Ischemic Optic Neuropathy Following Spine Surgery

Case Control Analysis and Systematic Review of the Literature

Anshit Goyal, MBBS; Mohamed Elminawy, MBBCh; Mohammed Ali Alvi, MBBS; Timothy R. Long, MD; John J. Chen, MD, PhD; Elizabeth Bradley, MD; Brett A. Freedman, MD; Mohamad Bydon, MD


Spine. 2019;44(15):1087-1096. 

In This Article


The incidence of perioperative ION has been estimated at 0.001% (1:61,000) in the general surgery population[13] while rates following spine surgery are estimated to be as high as 0.2% (2:1000).[12,14–17] This study represents the largest compilation of reports from the literature and institutional case series (n = 12). A matched case–control analysis was also performed using our own institutional cases to assess possible risk factors, in a more in-depth fashion. Patients from our institutional series were older in terms of mean age than patients from the literature review (67 yrs vs. 48 yrs, P < 0.001). Cases from our series had higher overall mean number of vascular risk factors per patient than patients from the literature review (2.3 vs.1.4, P < 0.003).

Consistent with other available reports of postoperative visual loss,[18,19] the majority of the cases were PION. Despite the availability of the postoperative visual loss registry by the ASA, individual case reports and case series remain the major source of detailed information regarding the condition. Due to medico-legal and other reasons, a significant number of cases are not reported.[3]

We confirmed that a majority of patients developing ION after spine surgery underwent prone positioning and had a lumbar or thoracolumbar procedure. A small but significant number of patients 6/58 (10.3%) developed ION despite a lack of prone positioning. In addition, our case–control analysis on our institutional cohort did not find prone positioning to be an independent risk factor for ischemic optic neuropathy. This might be attributed to the fact that while prone spine surgery may be a significant risk factor when compared with other non-ocular surgeries, it might lose significance when comparing different spine procedures, since the majority are performed in a prone position.

There is an interplay of a variety of risk factors playing an etiological role in development of ION. We found a significantly long operative duration, higher blood loss, and change in perioperative hemoglobin in these patients. Patients incurring ION also had higher number of operated vertebral levels and were more likely to have had a fusion. These findings were confirmed in our case–control analysis. Spinal fusions and vertebrectomies are especially characterized by higher blood loss. This, in conjunction, with prolonged prone positioning and intraoperative hypotension renders a higher risk for the development of ION (compared with decompressions and other non-spine surgeries). In longer procedures, the actual total duration of prone positioning also becomes important. Prolonged prone positioning may cause postoperative facial or periorbital edema thereby leading to indirect elevation of orbital venous pressures and contribute to ischemia.[20] In our review, facial edema was reported present in the majority of (24/41; 59%) cases in which facial edema was assessed. The incidence was 50% (6/12) within our own series. Also, 35/46 (76%) of cases were noted to have an "intraoperative hypotensive episode" defined as SBP of less than 100 mmHg for a duration of more than 2 minutes.[21] While deliberate hypotension is commonly utilized to control bleeding during surgery, prolonged hypotension has been noted as a risk factor of POVL in several published reports although its causative role is controversial.[3,12,22–24] In addition, in a review of 60,985 patients undergoing anesthesia, Roth et al[13] noted that overall duration of surgery was found to be an independent risk factor for eye injury. Patients undergoing cervical procedures formed the smallest subgroup in the overall cohort. This is consistent with one of the largest retrospective multi-institutional cohort studies by Gabel et al.[25] The authors analyzed 13,496 cases of cervical spine surgery from 21 high-volume surgical centers and did not find any patients with a complication of postoperative blindness within the 6-year study period.[20]

We also found that patients presenting with PION were likely to have worse visual acuity at initial baseline compared with patients with AION, although no differences in odds of improvement in vision were observed. Prospects for meaningful visual recovery remain bleak for both types of ION. While the usable data in this systematic review generated the conclusion that the rate of visual recovery is similar between AION and PION, it is worth noting that the subtype of ION was missing/unreported in an about a quarter of patients reported in the literature (n = 47/194) (24.2%).

It is important to note that the type of head support offered during positioning for the procedure might have a direct impact on the development of postoperative vision loss. The available options include foam pads, horseshoe headrests, and Mayfield pins or Gardner-Wells tongs. Direct pressure on the globe from horseshoe headrests has been attributed to the occurrence of vision loss due to central retinal artery occlusion (CRAO) and central retinal vein occlusion (CRVO) while its role in the development of ION remains unclear.[2,26,27]

From a management standpoint, cases presenting with new vision complaints following spine surgery should undergo an urgent ophthalmologic examination to ascertain the cause of vision loss. This examination will include testing of the pupillary reflex. Absence of this, implicates injury of the optic nerve (i.e., ION) or globe, while presence of a normal pupillary reflex indicates cortical blindness from either stroke or posterior reversible encephalopathy syndrome (PRES). For prolonged spinal procedures, it might be worthwhile to stage the surgery to reduce the duration and intensity of each operative stage. Deliberate hypotensive anesthesia must be used cautiously while avoiding prolonged periods of hypotension. Using soft headrests to avoid undue pressure on the globe during prone positioning combined with regular eye checks by the anesthesiologist during the procedure might decrease the risk of periorbital swelling.[20]

While there may be some factors that may increase the risk of ION, it remains worthwhile to note that it is a random catastrophic event that can occur even in patients in which all modifable risks are addressed and nonmodifable risk factors are absent, highlighting the need to adequately counsel patients of this rare but real risk of spinal surgery.

There were limitations to this review. First, we encountered missing data in our series which was inherent to its retrospective nature. Second, the information available from individual reports in the literature was highly heterogeneous, making quantitative synthesis difficult with variable denominators. While lack of data proved to be challenging, this highlights the need for active reporting and brings attention to this devastating complication. Fourth, our case–control analysis was limited by a small sample size, possibly rendering important risk factors statistically insignificant and limiting the ability to perform multivariable analysis. It is worthwhile to mention here, the largest multicenter case–control study (80 cases; 315 controls) by the Postoperative Visual Loss Study Group, which focused on assessing risk factors for ION following spinal fusion using patients from the ASA registry.[28] The authors found male sex, obesity, use of the Wilson frame, duration of anesthesia, estimated blood loss (EBL), and intraoperative blood pressure below 40% of baseline more than or equal to 30 minutes to be significantly associated with ION. Interestingly, while we found higher number of operated levels to be a risk factor, number of levels fused was not significantly associated with ION in their analysis. However, while the study included patients undergoing spinal fusion only, we included all patients undergoing any spine surgery with the aim to provide a comprehensive review of existing literature supplemented with a case–control analysis from our own institution.