Statistics from Altmetric.com
We report the case of an insulin dependent diabetic who suffered transient blindness as a result of a change of posture following vitrectomy surgery with injection of gas.
This 34 year old woman with longstanding diabetic disease, postural hypotension, and proliferative retinopathy had undergone argon laser panretinal photocoagulation (PRP) to both eyes. Her right eye was amblyopic and had visual acuity of counting fingers (CF), whereas the left eye, at best, had vision of 6/12 that was frequently impaired by recurrent vitreous haemorrhage. She underwent left vitrectomy and required 30% sulphur hexafluoride (SF6) gas to support an incidental small, inferior detachment related to a round hole. The patient postured supine overnight and remained under the care of the physicians.
The following day her vision dropped suddenly and severely on her way to the eye clinic. Visual acuity was no perception of light (NPL) in the operated eye and CF in the right eye. The left pupil was already dilated and anterior segment examination was satisfactory. The view of the fundus was limited owing to the presence of gas and residual haemorrhage but the retina was flat. The intraocular pressure (IOP) was 42 mm Hg and her blood pressure (BP) was 70/40. A decision to withdraw 0.5 ml of gas from the eye was made. While the patient was lying down for this procedure she immediately noted an improvement in vision to light perception. Her blood pressure improved in the supine position to 150/80. Following removal of gas the IOP fell to 20 mm Hg and the vision improved to counting fingers. Four weeks later, with disappearance of the gas her best corrected vision had improved to 6/18 and the retina remained flat.
Judging by the severity of the visual loss we believe that it was due to obstruction of both the central retinal artery and optic nerve perfusion. The precipitating factor in this case would appear to have been the change in posture from supine to erect. The presence of gas in the operated eye had raised the intraocular pressure to a level such that the perfusion pressure to the eye was effectively eliminated resulting in a period of no light perception. Restoring the patient to the supine position was sufficient alone to reverse the condition and improve the perfusion of the eye. When the IOP was reduced to normal she was out of this critical situation and was able to maintain the ocular perfusion, even when sitting.
Collapse of the central retinal artery is known to occur in conditions of high IOP1 but vision does not fluctuate with posture. However, it is known that raised IOP also reduces orthograde axonal transport and compromises nerve head perfusion.2
Postural hypotension in diabetics is secondary to autonomic neuropathy. The pathophysiology is not clear but it seems the main factor is blunted catecholamine response to standing and failure of the lower limb vascular resistance to increase adequately.3 Systemic hypotension can reduce optic nerve blood flow as illustrated by numerous reports in the literature of non-arteritic anterior ischaemic optic neuropathy (NA-AION) occurring following acute blood loss.4–6 Patients with NA-AION have a significantly greater postural pressure change in IOP compared with healthy subjects as shown by James and Smith.7 In the same study they demonstrated a higher pulsatile ocular blood flow (POBF) on standing compared with supine. In this patient with autonomic neuropathy it is likely that the normal increase in POBF on standing was inadequate.
We advise particular caution in diabetics undergoing ocular surgery in which there may be a significant postoperative pressure rise. Similar problems can occur during surgery if hypotensive anaesthesia is either deliberately or inadvertently employed in diabetics. Beware of assuming poor postoperative vision to be purely the result of the presence of haemorrhage or gas; patients may have quite poor vision immediately after vitreoretinal surgery but will rarely have NPL unless there is also optic nerve compromise. We also suggest one should to aim for a lower postoperative IOP in diabetics.