Statistics from Altmetric.com
Editor,—Diabetic retinopathy (DR), with its complications, is the leading cause of blindness among the working population in developed countries. Asymmetric diabetic retinopathy (DR) has been defined as proliferative disease in one eye and background or no retinopathy in the fellow eye, persisting for at least 2 years.1 Previous reports have described systemic and local factors associated with the development of asymmetric disease.1 2 These include unilateral carotid artery stenosis, chorioretinal scarring, complete posterior vitreous detachment, amblyopia, unilateral elevated intraocular pressure, optic atrophy, retinal pigment epithelial atrophy, myopia ⩾5D, anisometropia >1D, concurrent retinal vascular disease, cataract extraction, vitreous loss, trauma, radiation, tumour, and unilateral recurrent panuveitis.
We observed a patient with proliferative DR in the right eye and no proliferative changes in the left eye which had Fuchs’ heterochromic cyclitis (FHC). In the absence of other known risk or protective factors, FHC was felt to have protected against the development of proliferative DR. The significance of this new observation and the possible mechanisms are discussed.
The patient was a 56 year old insulin treated type II diabetic. He also had systemic hypertension and was a smoker. FHC of the left eye was diagnosed on the basis of typical stellate keratic precipitates scattered over the entire corneal endothelium, chronic low grade anterior uveitis, iris heterochromia, and posterior subcapsular cataract. Pharmacological testing with 4% cocaine excluded a diagnosis of Horner’s syndrome in the left eye.
He suffered widespread vascular complications of his combined diabetic and hypertensive state, including nephrotic syndrome and peripheral vascular disease culminating in left below knee amputation. These were accompanied by the development of new vessels at the disc in the right eye (Fig 1A), but only background changes in the left eye, although ischaemic changes were evident on fundus fluorescein angiography (Fig1B).
At his initial assessment in the ophthalmology clinic, best corrected Snellen visual acuities were right eye 6/6; left eye 6/9. The refractive errors were right eye +1.00/+0.50@180; left eye +2.00 DS. Intraocular pressures were 16 mm Hg either eye. Retinal pigment epithelial changes were present at the left macula (Fig 1B) but there was no extensive retinal pigment epithelial atrophy or chorioretinal scarring. There was neither proliferative diabetic retinopathy nor clinically significant macular oedema in either eye.
Nine months after the first examination he developed clinically significant macular oedema in the right eye. This was treated with focal argon laser with resolution of the macular oedema. Two years later new vessels at the right disc (Fig 1A) were noted and panretinal photocoagulation was performed (2707 burns in two sessions). Despite laser treatment, vitreous haemorrhage ensued. Further laser treatment to the right eye (1149 burns) led to regression of new vessels at the right disc (Fig 2A).Cataract progressed in the eye with FHC so that 5 years after his first visit to the ophthalmology department left phacoemulsification with heparin coated intraocular lens implant was required. Extracapsular cataract extraction is associated with accelerated retinopathy progression postoperatively.3Accelerated retinopathy was defined as the development of any of the following within 6 months postoperatively: new clinically significant macular oedema, recurrent clinically significant macular oedema in eyes that had preoperative resolution of macular oedema after focal laser treatment, increased hard exudates or intraretinal haemorrhages in eyes with clinically significant macular oedema at the baseline examination, or new onset proliferative diabetic retinopathy. No such progression occurred in this patient. Six months after left phacoemulsification with intraocular lens implantation there was no clinically significant macular oedema and best corrected visual acuity was 6/6. Despite evidence of worsening retinal ischaemia there was no progression to proliferative diabetic retinopathy (Fig 2B).
He was investigated for any evidence of local or systemic factors which may have contributed to the asymmetric retinopathy. There was no anisometropia. Intraocular pressures were equal in both eyes. There was no posterior vitreous detachment in either eye, no optic atrophy, and visual fields were full. Ultrasound of the carotids excluded haemodynamically significant stenosis. Retinal macrocirculation and microcirculation were assessed by fundus fluorescein angiography. Arm-retina times were right eye 11.1 seconds; left eye 11.0 seconds (normal 10.9 (SD 2.6) seconds).4
Kohner et al have put forward a working hypothesis for the pathogenesis of DR.5 6 The first change is hyperperfusion initiated by hyperglycaemia and influenced by high blood pressure and impaired autoregulation. The hyperglycaemia damages both pericytes and endothelial cells. The increased blood flow results in further damage to vessel walls, occlusion of some vessels, hypoxia, and ischaemia, resulting in proliferative DR. Factors which reduce or normalise retinal blood flow therefore have a protective effect in DR.
Although the aetiology of FHC is unknown, a vascular pathogenesis is one of the proposed hypotheses for its cause. An immune complex vasculitis may be the cause of abnormal hyalinisation of the iris vessel walls previously described.7 Ultimately there is narrowing of the vessel lumen, or even occlusion. This may explain the rubeosis and neovascular glaucoma sometimes seen in eyes with FHC.
Sympathetic theories for FHC have also been proposed, although a sympathetic aetiology has never been proved.8 Loewenfeld and Thompson9 felt there was inadequate evidence to support the proposed connection with sympathetic paralysis or denervation. Despite the arguments put forward in their review to reject the association between FHC and hemifacial atrophy (Parry-Romberg syndrome), and the hypothesis of a sympathetic defect implicated in both diseases, many authors still support this theory.10 Sympathicoparalysis explains the increased permeability of the blood-ocular barrier, with escape of cell elements, primarily albumin and lymphocytes, into the aqueous7 and vitreous. If denervation hypersensitivity occurs, because of an increase in receptor sites following destruction of postganglionic neurons, there may be rebound vasoconstriction with reduction in blood flow.
The arm-retina time measured by fluorescein angiography is a measure of the vascular system supplying the eye. The absence of any haemodynamically significant carotid artery stenosis and the normal arm-retina time suggest that there were no haemodynamic factors, even in the more distal branches of the internal carotid system, which contributed to the asymmetric retinopathy. In fact, pulsatile ocular blood flow in the left eye was normal (1101 μl/min), despite signs of ischaemia on fluorescein angiogram (Fig 2B). Ocular blood flow in the right eye was also normal (824 μl/min), although this was after extensive panretinal photocoagulation.6 Our hypothesis, therefore, is that FHC protected his left eye from progression to proliferative diabetic retinopathy.
This case suggests that FHC protected against proliferative DR but the mechanism is unclear and merits further consideration. Hopefully, this new observation will help to increase our understanding of these complex diseases and eventually affect the formulation of clinical practice.