Perioperative ischemic optic neuropathy (ION) is rare but often causes profound visual loss. Such patients frequently present to the neuro-ophthalmologist, sometimes with medicolegal issues related to causation. Two experts debate whether anemia and hypotension are causally related to perioperative ION.
Pro: Anemia and Hypotension Are Causally Related to Perioperative Ischemic Optic Neuropathy: Rod Foroozan, MD
Ischemic optic neuropathy (ION), the most common acute optic neuropathy in individuals aged 50 years and older, is often categorized by anatomic site. When the presumed vascular insufficiency involves the optic nerve anterior to the lamina cribrosa causing optic disc edema, it is referred to as anterior ischemic optic neuropathy (AION). When the presumed vascular disorder occurs posterior to the lamina cribrosa it is termed posterior ischemic optic neuropathy (PION), occurring without optic disc edema (1). ION has also been characterized by the underlying pathophysiology either from vasculitis (arteritic) or without the evidence of vasculitis (nonarteritic).
Nonarteritic ischemic optic neuropathy (NAION) is the most common form of ION with several risk factors cited, yet often without an obvious precipitating event in most patients (2). In other patients with ION, a triggering event has been identified, including acute hemorrhage (3) and the circumstances of the perioperative setting. Why some patients develop perioperative AION and others PION is unclear, but patient-specific factors may include small cup-to-disc ratio (4,5) variations in vascular anatomy (6) and autoregulatory dysfunction (7). For this discussion, AION and PION are lumped together as, apart from the presence of optic disc edema, there is little to distinguish them clinically.
Perioperative ION, defined as ION occurring relating to the operative or postoperative periods, is likely the most common cause of fixed visual loss after nonocular surgery (8). Visual symptoms from perioperative ION most commonly begin immediately after surgery but may be delayed because of the effects of anesthesia and the focus often paid to the main surgical procedure. In some patients with perioperative AION, there may be a delay in visual symptoms from the time of the onset of disc edema, like in typical spontaneous NAION (9). Most commonly, ischemic damage is limited to the optic nerve and, at least clinically, typically spares the remainder of the central nervous system. Patients are frequently told of the risk to life and other severe morbidity related to nonocular surgery; however, visual deficits are often omitted from the preoperative discussion between the surgeon and the patient.
Perioperative ION has been reported in relation to a wide variety of surgical procedures. Most often, these procedures require long operative times (often more than 6 hours), such as spine surgery, cardiopulmonary surgery, and neck dissection. The precise incidence of perioperative ION has been estimated to be 0.01% in patients undergoing spinal fusion (10) and 0.018% of patients undergoing cardiopulmonary bypass (11). A prevalence study based on a nationwide inpatient sample of multiple nonocular surgical procedures noted that perioperative visual loss (including ION) occurred in 8.64 of 10,000 patients undergoing cardiac surgery and 3.09 of 10,000 patients undergoing spinal fusion, as compared to 0.12 of 10,000 patients undergoing appendectomy. Because of the perceived scope of the problem in spinal fusion surgery, as of July 1, 1999, the American Society of Anesthesiologists Committee on Professional Liability began to compile data on postoperative visual complications, the Postoperative Visual Loss (POVL) Registry.
The pathogenesis of perioperative ION has remained unclear but is thought to occur as a result of hemodynamic derangements. These include a decrease in arterial perfusion pressure, an increase in resistance to blood flow, and a decrease in oxygen carrying capacity. A number of potential factors have been thought to play a role, including anemia, hypotension, venous pressure, head-down position during surgery, increased cerebrospinal fluid pressure, and direct ocular compression (Fig. 1). Several of these factors are difficult to quantitate, and most patients have more than 1 factor involved. It is clear that no one mechanism is responsible in every patient.
In part, because of the infrequent occurrence of perioperative ION, attempts to tease apart putative risk factors have been largely confined to case reports and epidemiologic reviews. The latter often lack all the clinical details of individual patients. Anemia and hypotension have been more commonly cited as risk factors because they are easily quantifiable. Anemia (from chronic anemia, intraoperative blood loss, and hemodilution) results in decreased oxygen carrying capacity of blood and hypotension (from hypovolemia and deliberate hypotension employed during surgery) causes decreased arterial perfusion pressure.
Accepting that Level I evidence is lacking, the effects of anemia and hypotension playing a role in perioperative ION can be summarized from the following case reports, epidemiologic studies, ocular procedures, and animal studies.
- Buono and Foroozan (12) reviewed 83 published patients to better characterize perioperative PION. The analysis considered both anemia and hypotension as putative risk factors. When the data were available, lowest reported perioperative hemoglobin was 9.5 g/dL (range: 5.8–14.2 g/dL) and the mean of the lowest reported hematocrit was 27.7% (range: 21%–40%). The mean drop in hemoglobin from the preoperative value to the lowest value perioperatively was 3.4 g/dL (range: 0.7–6.0 g/dL), and the mean drop in hematocrit from the preoperative value to the lowest perioperative value was 14.4% (range: 9.5%–24%). The mean intraoperative estimated blood loss (EBL) was 3.7 L (range: 0.8–16 L). Regarding hypotension, the mean of the lowest recorded perioperative systolic blood pressure (SBP) was 77 mm Hg (range: 48–120 mm Hg). The mean drop in SBP from the preoperative value to the lowest recorded perioperative value was 53 mm Hg (range: 22–85 mm Hg). Nearly without exception, the patients in this review had some degree of perioperative anemia and hypotension. Some degree of anemia and hypotension has been documented in the overwhelming majority of other case reports of perioperative ION (13–15) since the review by Buono and Foroozan (12).
- Additional support for the role of anemia and hypotension comes from epidemiologic studies from the POVL Registry. In a multicenter case–control design study, 80 patients with perioperative ION after spinal fusion were compared to 315 control subjects without perioperative ION. After multivariate analysis, risk factors for ION included male gender, obesity, Wilson frame use, anesthesia duration, EBL (odds ratio per liter was 1.34), and colloid as percent of nonblood replacement (16). A review of a nationwide hospital database of 2,511,073 patients who underwent spinal fusion noted that transfusion (with an incident rate ratio of 2.72 and presumably previous anemia prompting transfusion) was a risk for the development of perioperative ION (10).
- Although there are relatively few studies, perioperative ION has been reported with ocular surgical procedures. It may be that perioperative ION has not been reported as frequently in relation to ocular surgery because vision loss is a well-known and often discussed potential complication of surgery involving the eye. Procedures such as cataract surgery, which do not involve blood loss or require general anesthesia, are less likely to be associated with anemia and systemic hypotension (17); however, in a report of patients who underwent vitrectomy for retinal detachment, 7 patients developed visual loss, with clinical findings suggestive of an optic neuropathy (18). Four of these patients underwent general anesthesia without retrobulbar anesthesia and a PION mechanism was thought to be the cause of visual loss as no patient developed optic disc edema. Five (71%) of the 7 patients were noted to have decreased ocular perfusion pressure (calculated as the difference between the mean arterial blood pressure and the intraocular pressure) with associated systemic hypotension compared to 7 (17%) of 42 patients in the control group.
- The effects of anemia and hypotension have largely been inferred from measurement of the blood pressure and hemoglobin more peripherally (systemically) than in the milieu of the optic nerve. In an attempt to directly assess optic nerve blood flow, Lee et al (19) anesthetized pigs with isoflurane and subjected them to conditions of euvolemic or hypovolemic hypotension, anemia, and venous congestion. Fluorescent microspheres were injected via a left ventricular catheter to measure cerebral blood flow (CBF) and optic nerve blood flow, with a calculation of oxygen delivery. There was no significant change in CBF or optic nerve blood flow in pigs with euvolemic hypotension when compared to controls. In hypovolemic hypotension, there was stable CBF and cerebral oxygen delivery; however, oxygen delivery to the optic nerve was significantly reduced. Furthermore, there was a compensatory increase in CBF with hypovolemic anemia and stable cerebral oxygen delivery. However, in those with hypovolemic anemia, compensatory mechanisms to the optic nerve were not apparent with a significant reduction in optic nerve oxygen delivery. These changes are consistent with the clinical picture that is often seen in perioperative ION: a relatively selective involvement of the optic nerve with sparing of the remainder of the central nervous system. These results lend support not only to the role of anemia and hypotension in perioperative ION but also to the potential importance of autoregulation (Fig. 1).
There have been little data published regarding treatment of perioperative ION. No therapy has been reported to have a beneficial effect with the possible exception of correction of hemodynamic derangements including with blood transfusion. However, the natural history of perioperative ION includes visual recovery in some patients so that the visual improvement may have occurred spontaneously.
Some patients, including those with a prior history of typical ION, inquire about the potential for perioperative ION (20). Apart from a small optic disc cup that may predispose to perioperative AION, I tell patients and family members that I am not aware of any modifiable risk factors or preventative measures proven to prevent perioperative ION.
Con: Anemia and Hypotension Are Not Causally Related to Perioperative Ischemic Optic Neuropathy: Karl C. Golnik, MD, MEd
Perioperative ION is the most common cause of perioperative vision loss (21). As Dr Foroozan has described, the 2 types of perioperative ION (anterior and posterior) will be considered together. It most frequently occurs after spine surgery in the prone position and coronary artery bypass grafting (22). Reported incidences range from 0.017% to 0.36% after spine surgery (23–26) and as high as 0.33% after coronary artery bypass grafting (23).
A variety of potential risk factors have been proposed for perioperative ION, including anemia, blood transfusion, hypotension, facial/orbital edema, vasopressors, male gender, Wilson frame use, prone surgical position, long anesthetic time, significant blood loss, and decreased percent colloid administration (26,27). Pre-existing comorbidities, such as obesity, diabetes mellitus, hypertension, smoking, and heart disease, also may be relevant (27,28). Most reports represent low levels of evidence (Level V) and, thus, controversy remains as to the exact mechanism(s), leading to perioperative ION. Fortunately, several fairly recent case–control and cohort studies give somewhat better evidence (Level IV) (23,24,27).
Patil et al (24) conducted a retrospective cohort study using the National Inpatient Sample to investigate perioperative ION after spine surgery. All patients from 1993 to 2002 undergoing spine surgery and who were documented as having vision loss were included. Univariate and multivariate analyses were used to determine association between visual loss after spine surgery and age, gender, diabetes, hypertension, hypotension, anemia, and blood transfusion. Duration of surgery could not be evaluated because the database does not include this information. Perioperative-ischemic optic neuropathy (POP-ION) was reported in 271 patients and anemia, age (65–84 years more likely than younger), hypotension, peripheral vascular disease, diabetes, and prone position were predictors of POP-ION. In addition, Myers et al (27) reviewed 37 cases of POP-ION and noted that patients with similar degrees of anemia and hypotension often do not have POP-ION (27).
Holy et al (23) performed a retrospective case–control analysis of 17 patients (of 126,666 operations from 1998 to 2004) who developed perioperative ION after a variety of nonocular surgeries. Each of the 17 patients was matched with 2 control patients who did not lose vision after surgery. Multiple preoperative, intraoperative, and postoperative variables (including both anemia and hypotension) were assessed and none were found to increase the risk of perioperative ION.
The Postoperative Visual Loss Study Group conducted a multicenter case–control review comparing 80 patients with perioperative ION after spinal surgery (between 1991 and 2006) to 315 control subjects without perioperative ION after spinal surgery (16). Potential risk factors included age, gender, hypertension, diabetes, smoking, atherosclerosis, and obesity. Other factors examined included fusion location (lumbar vs nonlumbar), indication for surgery, and blood pressure. Procedural factors included type of surgical frame, number of levels of fusion, and type of headrest. Intraoperative factors included anesthetic duration, EBL, decrease in blood pressure, lowest hematocrit, fluid management variables, and use of vasopressors. Stepwise multivariate analysis showed obesity, male gender, Wilson frame, longer anesthetic time, greater EBL, and lower percent colloid administration to be independent risk factors for development of perioperative ION. Anemia and hypotension were not independent risk factors. The authors comment: “The lack of an independent effect of anemia or any blood pressure more than 40% below baseline for 30 min in the multivariate analysis demonstrates the importance of using detailed perioperative data on control subjects to assess whether or not the effect of these factors remains significant when other relevant intraoperative data such as anesthesia duration, EBL, and volume administration are analyzed.” The Study Group recommended keeping head above heart level, minimizing prone position and good hemostasis.
The American Society of Anesthesiologists Task Force on Perioperative Vision Loss has issued 2 Practice Advisories for spine surgery; the first in 2006 and an updated version in 2012 (29,30). The task force concluded that there are no identifiable preoperative risk factors for perioperative ION. They do acknowledge that prolonged procedures and substantial blood loss may increase the risk of perioperative ION. Deliberate hypotentive techniques did not predispose to vision loss, and although hemoglobin and hematocrit should be monitored, there is no documented lower hemoglobin limit that has been associated with perioperative ION.
In summary, the data regarding risk factors are poor, much of it being Level V. The data summarized from Level IV studies indicate that anemia and hypotension are not independent risk factors for perioperative ION. I do not believe the data from the Postoperative Visual Loss Study Group were available for the most recent American Society of Anesthesiologists Task Force on Perioperative Vision Loss advisory. It will be interesting to see how these new data and needed future studies will shape the next Practice Advisory.
Rebuttal: Dr. Foroozan, MD
The literature suggests that anemia and hypotension are overwhelmingly important but not the only factors involved in perioperative ION. However, perioperative ION is distinctly rare in the absence of anemia and hypotension.
Dr. Golnik correctly points out that the analysis of the POVL Registry did not show that anemia and hypotension are independent risk factors for the development of perioperative ION. The problem with this analysis is that it has only been in patients undergoing spine surgery so that the generalizability is limited (31). By its epidemiologic nature, this analysis cannot account for all the individual factors involved for each patient. Grouping patients together as “cases” and “controls” cannot account, for example, for the potential of relative hypotension and anemia for a particular patient. The decline in blood pressure and hemoglobin may be sufficient to cause ischemia in one patient, although in another with similar findings, optic neuropathy may not occur.
The single institution study of Holy et al (23), documenting perioperative ION in 17 patients, demonstrated that intraoperatively the mean arterial blood pressure decreased by 41.8% and the hematocrit fell by 41.6% in patients with visual loss. Although not statistically different from the control group, anemia and hypotension were clearly present in patients who developed perioperative ION.
It is tempting to try and find a single unifying answer for why perioperative ION occurs; however, as this has appeared increasingly less likely, the clinical characteristics of each patient must be weighed in light of the particular type of surgical procedure. Hopefully a reliable, noninvasive method to assess the oxygen carrying capacity, perfusion pressure, and resistance to blood flow at the level of the optic nerve will be developed. This coupled with improved techniques to monitor visual function in the perioperative period will help better define which factors are most important in a particular patient with perioperative ION.
What seems most clear is that unanswered questions, which have been thought to be related to individual patient characteristics (question marks in Fig. 1), are clearly important in the pathophysiology of perioperative ION.
Rebuttal: Dr. Golnik, MD, MEd
Dr. Foroozan cites the publication of Rubin et al (10) noting transfusion to be a risk factor for perioperative ION. Dr. Foroozan opines that prior anemia prompted transfusion, which makes sense. However, similar to the other best data (23,27), anemia was not identified as a risk factor in this study either. Rubin et al (10) suggest that transfusion maybe a surrogate for blood loss, but this is different than anemia.
Dr. Foroozan cites a porcine model of optic nerve blood flow (19). The results are interesting and potentially relevant but it is unclear how applicable this model is to humans, where multiple risk factors have been identified (aging, obesity, etc) and cannot be replicated in the animal model.
I agree with Dr. Foorozan that “…the precise pathogenesis of perioperative ION has remained unclear.” Although it would seem to make sense that anemia and hypotension are in some way related to postoperative ischemic optic neuropathy, the best data do not support a causal relationship.
Conclusion: Andrew G. Lee, MD, and Greg Van Stavern, MD
Perioperative ION typically results in devastating visual loss and frequently has medicolegal implications. Drs. Foroozan and Golnik highlight the challenge of applying the available literature to the individual patient. It is reasonable to conclude that many cases of perioperative ION are multifactorial and that specific risk factors may vary in importance from patient to patient. In addition, treatment of one risk factor (e.g., hypotension) might increase the risk of another (e.g., intraoperative bleeding). The treatment of anemia (e.g., with transfusion) and the treatment of hypotension (e.g., with vasopressors) have risks (e.g., transfusion reaction, vasoconstriction) and paradoxically might increase rather than decrease the risk of POVL. Carefully evaluating potential risk factors in each individual patient and excluding alternate causes may be the best approaches for the time being. Discussion of the potential surgical risks during informed consent, particularly for spinal and cardiovascular surgical procedures, should include permanent visual loss.
1. Biousse V, Newman NJ. Ischemic optic neuropathies. N Engl J Med. 2015;372:2428–2436.
2. Arnold AC. Pathogenesis of nonarteritic anterior ischemic optic neuropathy. J Neuroophthalmol. 2003;23:157–163.
3. Hayreh SS. Anterior ischemic optic neuropathy. VIII. Clinical features and pathogenesis of post-hemorrhagic amaurosis. Ophthalmology. 1987;94:1488–1502.
4. Foroozan R. Shock-induced anterior ischaemic optic neuropathy after radical prostatectomy. Clin Exp Ophthalmol. 2004;32:438–439.
5. Foroozan R, Buono LM, Savino PJ. Optic disc structure and shock-induced anterior ischemic optic neuropathy. Ophthalmology. 2003;110:327–331.
6. Hayreh SS. Inter-individual variation in blood supply of the optic nerve head. Its importance in various ischemic disorders of the optic nerve head, and glaucoma, low-tension glaucoma and allied disorders. Doc Ophthalmol. 1985;59:217–246.
7. Prada D, Harris A, Guidoboni G, Siesky B, Huang AM, Arciero J. Autoregulation and neurovascular coupling in the optic nerve head. Surv Ophthalmol. 2016;61:164–186.
8. Newman NJ. Perioperative visual loss after nonocular surgeries. Am J Ophthalmol. 2008;145:604–610.
9. Hayreh SS. Anterior ischemic optic neuropathy. V. Optic disc edema an early sign. Arch Ophthalmol. 1981;99:1030–1040.
10. Rubin DS, Parakati I, Lee LA, Moss HE, Joslin CE, Roth S. Perioperative visual loss in spine fusion surgery: ischemic optic neuropathy in the United States from 1998 to 2012 in the Nationwide Inpatient Sample. Anesthesiology. 2016;125:457–464.
11. Nuttall GA, Garrity JA, Dearani JA, Abel MD, Schroeder DR, Mullany CJ. Risk factors for ischemic optic neuropathy after cardiopulmonary bypass: a matched case/control study. Anesth Analg. 2001;93:1410–1416.
12. Buono LM, Foroozan R. Perioperative posterior ischemic optic neuropathy: review of the literature. Surv Ophthalmol. 2005;50:15–26.
13. Chang SH, Miller NR. The incidence of vision loss due to perioperative ischemic optic neuropathy associated with spine surgery: the Johns Hopkins Hospital Experience. Spine (Phila Pa 1976). 2005;30:1299–1302.
14. Battu R, Prasad A, Kanchi M. Perioperative optic neuropathy in patients undergoing off-pump coronary artery bypass graft surgery. Ann Card Anaesth. 2014;17:92–97.
15. Price TP, Ivashchenko A, Schurr MJ. Perioperative visual loss after excision and autografting of a thermal burn to the back. Burns. 2014;40:e31–e34.
16. Postoperative Visual Loss Study Group. Risk factors associated with ischemic optic neuropathy after spinal fusion surgery. Anesthesiology. 2012;116:15–24.
17. McCulley TJ, Lam BL, Feuer WJ. A comparison of risk factors for postoperative and spontaneous nonarteritic anterior ischemic optic neuropathy. J Neuroophthalmol. 2005;25:22–24.
18. Bansal AS, Hsu J, Garg SJ, Sivalingham A, Vander JF, Moster M, Macguire JI, Regillo CD. Optic neuropathy after vitrectomy for retinal detachment: clinical features and analysis of risk factors. Ophthalmology. 2012;119:2364–2370.
19. Lee LA, Deem S, Glenny RW, Townsend I, Moulding J, An D, Treggiari MM, Lam A. Effects of anemia and hypotension on porcine optic nerve blood flow and oxygen delivery. Anesthesiology. 2008;108:864–872.
20. Lee MS, Armbrust KR, Foroozan R, Pomeranz HD. Perioperative visual loss in ocular and nonocular surgery. Am Acad Ophthalmol. 2012;10:1–14.
21. Ho VT, Newman NJ, Song S, Ksiazek S, Roth S. Ischemic optic neuropathy following spine surgery. J Neurosurg Anesthesiol. 2005;17:38–44.
22. Roth S. Postoperative Blindness, Anesthesia. 6th edition. Miller RD, ed. New York, NY: Elsevier/Churchill Livingstone, 2005:2991–3020.
23. Holy SE, Tsai JH, McAllister RK, Smith KH. Perioperative ischemic optic neuropathy: a case control analysis of 126,666 surgical procedures at a single institution. Anesthesiology. 2009;110:246–253.
24. Patil CG, Lad EM, Lad SP, Ho C, Boakye M. Visual loss after spine surgery: a population-based study. Spine (Philia Pa 1976). 2008;33:1491–1496.
25. Shen Y, Drum M, Roth S. The prevalence of perioperative visual loss in the United States: a 10-year study from 1996 to 2005 of spinal, orthopedic, cardiac, and general surgery. Anesth Analg. 2009;109:1534–1545.
26. Stevens WR, Glazer PA, Kelley SD, Lietman TM, Bradford DS. Ophthalmic complications after spinal surgery. Spine (Philia Pa 1976). 1997;22:1319–1324.
27. Myers MA, Hamilton SR, Bogosian AJ, Smith CH, Wagner TA. Visual loss as a complication ofspine surgery. A review of 37 cases. Spine (Philia Pa 1976). 1997;22:1325–1329.
28. Katz DM, Trobe JD, Cornblath WT, Kline LB. Ischemic optic neuropathy after lumbar spine surgery. Arch Ophthalmol. 1994;112:925–931.
29. American Society of Anesthesiologists Task Force on Perioperative Blindness. Practice advisory for perioperative visual loss associated with spine surgery: a report by the American Society of Anesthesiologists Task Force on Perioperative Blindness. Anesthesiology. 2006;104:1319–1328.
30. American Society of Anesthesiologists Task Force on Perioperative Vision Loss. Practice advisory for perioperative visual loss associated with spine surgery: an updated report by the American Society of Anesthesiologists Task Force on Perioperative Blindness. Anesthesiology. 2012;116:274–285.
31. Duggal R, Menkes DB. Evidence-based medicine in practice. Int J Clin Pract. 2011;65:639–644.