Blindness is a dreaded complication of spinal surgery. Fortunately, it is very rare, with an estimated prevalence of 1/2,000 to 1/100,000. Rare complications are difficult to study at a single institution due to the low numbers involved, so meta-analysis and registry data can be helpful to generate a sufficient number of cases to allow for an analysis of risk factors. Dr. Goyal and colleagues from the Mayo Clinic performed an interesting hybrid study that included a case-control study of 12 patients with perioperative visual loss from their institution and a systematic review of the currently available literature. They matched the 12 cases with 48 age and sex matched controls and then performed a logistic regression to analyze potential risk factors. The average age of their cases was 66, and 77% were male. Eleven of the twelve patients underwent a fusion, and all but one had surgery involving at least three levels. None had isolated cervical surgery, and 11/12 underwent surgery in the prone position. The mean blood loss was 1.5 L (range 100 mL-10 L), and the mean operative time was 561 minutes (over 9 hours). Five patients experienced blindness, with three not improving at follow-up and two lost to follow-up. The other 7 had either a severe or mild deficit initially, and none had a major visual loss at follow-up. The majority (10/12) were classified as posterior ischemic optic neuropathy, and 75% were bilateral. The logistic regression analysis revealed that fusion, increasing number of levels treated, blood loss, volume of intra-operative crystalloid infused, and change in hematocrit were all independent predictors of perioperative visual loss. In the systematic review, they identified 182 patients in case reports and case series, the largest of which included 83 patients from the American Society of Anesthesiologists (ASA) perioperative visual loss registry. The systematic review patients were younger (average age 48), and 81% underwent multilevel surgery. The average blood loss in the literature review patients was over 2 L (range 100 mL to 25 L), and the mean operative duration was 493 minutes (over 8 hours).
The authors have published a very thorough study including 12 cases from their home institution and 182 patients from their literature review. This likely represents the largest number of perioperative visual loss patients ever assembled. The study does not offer any surprising new conclusions, but it does confirm our understanding that longer surgery with greater blood loss increases the risk. At the same time, the complication is somewhat random, and cases were identified with single level surgery and 100 mL of blood loss. While extremely rare, cases with supine positioning were also reported. The authors did not identify any modifiable risk factors, so it is not clear if any interventions can decrease the risk. They simply concluded that patients need to be educated about the risk and consider it when making the decision to undergo spine surgery. While the literature review yielded a relatively high number of patients, the analyses that could be performed on that cohort were very limited due to the variable reporting of risk factors and outcomes. As such, the risk factor analysis included only the 12 cases from the authors' institution, which limited the statistical power. They also did not analyze the type of head positioner used (i.e. foam cushion, horseshoe, Mayfield tongs), which could play a role. The authors also controlled for age and gender, which may have been risk factors. Given that the ASA registry included only 83 patients from fifteen years of data collection, it is clear that assembling a sufficient number of cases to study this complication is very difficult. All patients undergoing spine surgery should be educated about the risk of perioperative visual loss, and those undergoing major surgery should understand that their risk is higher (though still quite low in absolute terms).
Please read Dr. Goyal's paper on this topic in the August 1 issue. Does this change your view on perioperative visual loss associated with spine surgery? Let us know by leaving a comment on The Spine Blog.
Adam Pearson, MD, MS
Associate Web Editor