Postoperative loss of vision is a devastating complication. It may be due to central retinal artery occlusion or to postoperative ischaemic optic neuropathy. We report here four cases of loss of vision.
A 74-year-old female patient underwent a total knee arthroplasty second revision. Her medical history included arterial hypertension controlled by diuretic treatment. Routine analysis was normal. General anaesthesia was induced with fentanyl, midazolam, thiopental and atracurium and maintained with isoflurane in O2 and N2O. Surgery lasted 3 h. Fluid loading was 1 l of Hartman's solution and 1 l of hetastarch. A blood salvage device allowed treatment of 3.7 l and restitution of 1 l. Intraoperative systolic blood pressure (BP) was maintained between 80 and 120 mmHg. Wound blood drainage amounted to 0.8 l during the first 24 h, and 0.5 l of hetastarch was administered. From the evening of day 1, the patient complained of blurred vision that reached complete blindness at day 3. Haemoglobin level was 6.6 g dl−1. Four units of packed red cells were transfused. The ophthalmologist noted absence of light perception. Fundal examination showed disc swelling with thin, fine and rigid arteries. Fluorescein retinal angiography confirmed bilateral anterior ischaemic optic neuropathy (ION) in an atheromatous patient. No recovery occurred.
A 49-year-old female patient with an unremarkable medical history underwent an arthroscopic meniscectomy. Spinal anaesthesia was performed with 40 mg hyperbaric lidocaine. An ephedrine infusion was simultaneously initiated (60 mg in 500 ml) with BP stable (120/70 mmHg). Fifteen minutes after tourniquet inflation and skin incision, the patient complained of sickness and anxiety. Systolic BP was 80 mmHg. The ephedrine infusion was accelerated and systolic BP increased to 180 mmHg. The hypotension had lasted less than 10 min. The day after, the patient complained of blurred vision. Fundal examination showed bilateral macular oedema. Fluorescein retinal angiography demonstrated transient retinal ischaemia, secondarily aggravated by ephedrine-induced vasospasm. Four years later, the patient still has a bilateral central scotoma with a visual acuity of 9/10 and 10/10 and a restriction on driving a car.
A 65-year-old black male patient underwent L2–L3 laminectomy and L2–L5 rearthrodesis. His medical history included hypertension treated by verapamil, hypercholesterolaemia, hyperuricaemia and renal insufficiency (blood creatinine 123 mmol l−1). General anaesthesia was induced with thiopental, midazolam, sufentanil and pancuronium. The patient was installed in the prone position with his head on a horseshoe headrest. Deliberate hypotension was achieved with isoflurane and isosorbide dinitrate (systolic BP 80–70 mmHg). Surgery lasted 6 h with 2.1 l blood loss. Two autologous packed red cell and fresh frozen plasma units were given. The intraoperative haematocrit was above 33%. The patient was extubated 12 h after surgery and immediately complained of complete blindness. Fundoscopy and retinal angiography were normal. Haemoglobin electrophoresis demonstrated heterozygous sickle cell trait and alpha-thalassaemia minor. A later fundoscopy showed normal papilla with a cherry red macula. The final diagnosis was a possible anterior ION, secondarily complicated by bilateral central retinal artery occlusion (CRAO).
A 54-year-old male patient was scheduled for subtotal glossectomy. His medical history included chronic abuse of alcohol and tobacco, diabetes mellitus requiring insulin, hypertension and one episode of acute pancreatitis and chemotherapy with cisplatyl and 5-fluorouracil. Anaesthesia was induced with propofol, midazolam and succinylcholine and maintained with remifentanil and desflurane in air. Surgery lasted 9 h and included bilateral radical neck dissection with right internal jugular vein section. Total fluid loading was 6 l of isotonic saline, 1 l of gelatin and 0.5 l of Hartman's solution. Intraoperative haemoglobin was 12.3–10.2 g dl−1. At the end of surgery, haemoglobin was 7.2 g dl−1, PaCO2 45 mmHg and pH 7.21, and major facial oedema was present that took 24 h to resolve enough to allow eye opening. Bilateral blindness was diagnosed at D1. Fundoscopy was normal, posterior ION was the final diagnosis.
Out of these four cases of permanent loss of vision (LOV), one might be considered as due to retinal ischaemia (transient retinal ischaemia due to hypotension and aggravated by ephedrine-induced vasospasm), two appeared to be due to ION, one of the anterior variety and one posterior, and one appeared to be an uncommon combination of initial ION secondarily complicated by CRAO.
The incidence of postoperative LOV is estimated to be between 0 and 0.2% [1–3]. It is thought to be due to a transient decrease in oxygen delivery to a watershed zone of the optic nerve or retina. In anaesthetized pigs submitted to specific conditions such as hypovolaemia, hypotension or anaemia compensatory mechanisms are able to maintain blood flow and oxygen delivery to the brain but not to the optic nerve . Two varieties of ION can be distinguished anatomically. Anterior ION (AION) concerns the optic nerve head, which is located anterior to the lamina cribrosa, the blood supply of which depends mainly on small branches of short ciliary arteries, which are end arteries with watershed zones. Raised intraocular pressure and globe compression may contribute to a decrease in perfusion pressure. The posterior variety (PION) affects the retrobulbar portion of the optic nerve, whose blood supply depends only on centripetal pia arteries. In both cases, swelling due to either ischaemia or venous hypertension may aggravate hypoperfusion. ION classically is bilateral and painless . Cerebral imaging may be normal or show optic nerve enlargement due to swelling. The main difference on fundoscopy is the presence of early disc swelling in AION compared with a normal fundoscopy in PION. Secondary optic disc pallor develops in both. CRAO is generally due to decreased retinal perfusion pressure due to arterial hypotension or intraocular hypertension. The most frequent is direct pressure on the globe from headrest pads. That is why CRAO is mostly unilateral . CRAO is easily diagnosed on fundoscopy. Differential diagnosis includes cortical blindness and ophthalmic venous obstruction.
Treatment of established LOV is disappointing. Transfusion, high-dose steroids, osmotic diuretics and even decompressive surgery for posterior ION have not proven to be efficacious .
Several risk factors for LOV have been identified . Many of them may be linked to hypotension and anaemia, which are considered to be central mechanisms: anaemia is frequently observed, as in cases 1 and 4 [1,5,7,8]. Hypotension is frequently associated, as in cases 1, 2 and 3 [5,8]. Excessive intraoperative bleeding or need for transfusion are probably linked to the previous  cases. High-volume crystalloid loading is also a risk factor, as in case 4 . It can act per se via an increase in central venous pressure or decrease in oncotic pressure to increase interstitial swelling that might aggravate optic nerve ischaemia. Facial swelling and chemosis are probably signs of such tissue oedema, as in case 4 . The prone position aggravated by associated head-down posture is an independent risk factor [1,9]. This might be due to an increase in venous pressure or intraocular pressure (IOP) or both. An increase in IOP (aggravated by the long duration of a procedure in the prone position) is a risk factor [1,10]. Finally, the role of vasopressor drugs has been suggested , as in case 3. In some patients, none of these specific risk factors were observed . The American Society of Anesthesiologists task force assumes that either prolonged procedures (>6.5 h) and substantial blood loss (>44% of blood volume) or both are present in the majority of cases after spinal surgery .
Procedures at risk for LOV include cardiac surgery, lumbar spinal surgery and radical neck dissection . Spinal surgery often requires a prone position and a head-down posture [1,8]. Radical neck surgery frequently involves jugular vein ligation that can increase ophthalmic venous pressure . Other procedures have been less frequently implicated such as liposuction, caesarean section, vascular or abdominal procedures. To our knowledge, we present here the first case after meniscectomy.
Patients' individual susceptibility may also be implicated, with a possible link to arterial disease , and a possible implication of hypertension, obesity, diabetes, tobacco abuse and atheroma. [1,5,6,8,12]. Case 4 appeared to have sickle cell syndrome. Finally, anatomical features such as a small disc-to-cup ratio or specific watershed zones may not be detected by routine assessment.
Maximal effort should be focused on LOV prevention. The American Society of Anesthesiologists task force advises mainly screening ‘at-risk patients’ wherever procedures involving prolonged duration and substantial blood loss are anticipated. Raising the haemoglobin threshold for transfusion should be considered. The need for prompt treatment of hypotension should make one reconsider the use of deliberate hypotension. Limiting crystalloid volume loading should make one consider colloid use. Posture must be carefully checked; and in the prone patient, the Trendelenburg posture should be replaced by a slight head-up position. Care must be taken to avoid direct compression of the eyeball. The horseshoe headrest should probably be replaced by a specific frame, such as Mayfield's pins, that avoids direct eyeball pressure and also jugular compression due to excessive head rotation.
1 Chang SH, Miller NR. The incidence of vision loss due to perioperative ischemic optic neuropathy associated with spine surgery: the John Hopkins Hospital Experience. Spine 2005; 30:1299–1302.
2 Warner ME, Warner MA, Garrity JA, et al
. The frequency of perioperative vision loss. Anesth Analg 2001; 93:1417–1421.
3 Newman NJ. Perioperative loss of vision after nonocular surgeries. Am J Ophthalmol 2008; 145:604–610.
4 Lee LA, Deem S, Gleeny RW, et al
. Effects of anemia and hypotension on porcine optic nerve blood flow and oxygen delivery. Anaesthesiology 2008; 108:864–872.
5 Lee LA, Roth S, Posner KL, et al
. The American Society of Anesthesiologists postoperative visual loss registry. Analysis of 93 spine surgery cases with postoperative visual loss. Anesthesiology 2006; 105:641–642.
6 ASA Task Force on Perioperative Blindness. Practice advisory for perioperative visual loss associated with spine surgery. Anesthesiology
7 Nuttal GA, Garrity JA, Dearani JA, et al
. Risk factors for ischemic optic neuropathy after cardiopulmonary bypass: a matched case/control study. Anesth Analg 2001; 93:1410–1416.
8 Dunker S, Hsu HY, Sebag J, et al
. Perioperative risk factors for posterior ischemic neuropathy. J Am Coll Surg 2002; 194:705–710.
9 Hunt K, Bajekal R, Calder I, et al
. Changes in intraocular pressure in anesthetized prone patients. J Neurosurg Anesthesiol 2004; 16:287–290.
10 Ozcan MS, Praetel C, Bhatti MT, et al
. The effect of body inclination during prone positioning on intraocular pressure in awake volunteers: a comparison of two operating tables. Anesth Analg 2004; 99:1152–1158.
11 Lee LA, Lam AM. Unilateral blindness after prone lumbar spine surgery. Anesthesiology 2001; 95:793–795.
12 Dilger JA, Tezlaff JE, Bell GR, et al
. Ischaemic optic neuropathy after spinal fusion. Can J Anaesth 1998; 45:63–66.