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


Case Series

Following Institutional Review Board approval, we retrospectively reviewed the hospital records of patients diagnosed with perioperative vision loss following spinal surgery at our institution between 2000 and 2017. Only cases in which the etiology of vision loss was attributed to either AION or PION were included in our analysis. The following information was abstracted from the medical records: demographics, comorbidities, perioperative variables such as type of spine surgery, positioning and approach and postoperative ophthalmologic examination. The severity of visual disturbance was classified according to the World Health Organization (WHO) criteria of visual acuity.[11]

Electronic Literature Search

A systematic search of Medline, Embase, Scopus was conducted by a master's level librarian with extensive electronic literature search experience through the Mayo Clinic library services from 1964 through September 2017 using Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines with the following keywords used as MeSH terms in all logical permutations: spine surgery and/or disease, ischemic optic neuropathy and/or visual, neurosurgical procedures, laminectomy, laminoplasty, spinal fusion, scoliosis, diskectomy, vertebral, intervertebral, optic nerve diseases or injuries, prone, preoperative, perioperative, postoperative. Figure 1 illustrates the search strategy. The results of the search were then screened by two authors (A.G. and M.E.) to determine eligibility for inclusion in the final review.

Figure 1.

PRISMA search strategy. PRISMA indicates Preferred Reporting Items for Systematic Reviews and Meta-Analyses.

Inclusion and Exclusion Criteria

We included case reports, case series, and registries reporting cases of ischemic optic neuropathy (ION) following spine surgery. References for included articles were also systematically reviewed to identify missed reports. We excluded duplicate studies, studies with uncertain final diagnosis, review articles with no patient level data, book chapters, abstracts, central retinal artery occlusion cases (CRAO), cases diagnosed with giant cell arteritis (GCA), and cortical blindness. Only English articles reporting human subjects were included. Given the nature of observational studies, the amount of data available was variable between different case reports. We excluded those case reports and case series which documented neither the postoperative or follow-up ophthalmic examination or overall visual outcome.

Case–control Analysis

For every case diagnosed at our institution, four controls undergoing spine surgery were randomly obtained by matching for age and year of surgery from a prospective institutional registry. Preoperative and perioperative risk factors for ischemic optic neuropathy were then compared between the two groups using conditional logistic regression.

Data Analysis

Descriptive statistical analysis was performed using JMP®, Version <10> SAS Institute Inc., Cary, NC, 1989 to 2007. We calculated pooled means and standard deviation for continuous variables. Categorical variables are reported as frequencies and proportions, we had variable denominator size due to heterogeneous data reporting in the literature. Pooling data from our institutional series and the literature, we also performed univariate logistic regression fitted for (i) blindness (Grade IV visual acuity) at initial presentation and (ii) overall improvement in vision to identify predictive factors. Significant risk factors were then used in a multivariable model controlling for patient age and sex. Case–control matching and conditional logistic regression was performed using R statistical software version 3.4.1. ( using the "optmatch" and "survival" packages, respectively.