Accepted Manuscript

Perioperative Visual Loss

Koffi M. Kla, M.D., Assistant Professor, Lorri A. Lee, M.D., Professor

PII: S1521-6896(15)00099-3
DOI: 10.1016/j.bpa.2015.11.004
Reference: YBEAN 881

To appear in: Best Practice & Research Clinical Anaesthesiology

Received Date: 19 October 2015

Revised Date: 2 November 2015
Accepted Date: 18 November 2015

Please cite this article as: Kla KM, Lee LA, Perioperative Visual Loss, Best Practice & Research Clinical
Anaesthesiology (2015), doi: 10.1016/j.bpa.2015.11.004.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT

Perioperative Visual Loss

Koffi M. Kla, M.D.

Assistant Professor

PT
Department of Anesthesiology

RI
Vanderbilt University Medical Center

1301 Medical Center Drive, 4648 TVC

SC
Nashville, TN 37232-5614

Email address: koffi.m.kla@vanderbilt.edu

Phone: 615-343-9419

U
AN
Fax: 615-936-6493
M
D

Corresponding Author

Lorri A. Lee, M.D.

TE

Professor
EP

Departments of Anesthesiology and Neurological Surgery

Vanderbilt University Medical Center

C

1301 Medical Center Drive, 4648 TVC

AC

Nashville, TN 37232-5614

Email address: lorri.a.lee@vanderbilt.edu

Phone: 615-343-9419

Fax: 615-936-6493
 ACCEPTED MANUSCRIPT

Abstract

Perioperative visual loss is an infrequent, devastating complication associated with spine surgery, most

commonly from ischemic optic neuropathy. Current research and expert opinion indicate that it is

associated with procedures that create elevated venous pressure in the head for prolonged periods of

PT
time. The largest case-control study on ischemic optic neuropathy associated with spine surgery found

RI
six independent and significant risk factors including male sex, obesity, Wilson frame use, longer

operative times, greater blood loss and a lower colloid to crystalloid ratio in the non-blood fluid

SC
administration. The American Society of Anesthesiologists developed a practice advisory for prevention

of this complication. Avoiding significant physiologic and hemodynamic perturbations as much as

U
possible in this setting is advisable given the uncertainty of the pathophysiology. Because prevention of
AN
this complication cannot be guaranteed, consent for perioperative visual loss should be strongly

considered for patients at high risk for this complication.

M
D

Key Words

Perioperative visual loss

TE

Ischemic optic neuropathy

EP

Spinal fusion

American Society of Anesthesiologists Postoperative Visual Loss Registry

C

Anesthesia
AC
 ACCEPTED MANUSCRIPT

Perioperative Visual Loss

Perioperative visual loss (POVL) is an uncommon but devastating complication related to spinal

surgery. As most patients undergoing spinal surgery are already debilitated by their spinal disease,

adding visual impairment can greatly decrease their productivity and quality of life. The reported

PT
incidence of POVL after spine surgery ranges from 0.03% to 0.2%.[1-3] Numerous factors have been

RI
proposed to contribute to the development of POVL including anemia, emboli, hypotension, globe

compression, prone positioning, volume and/or type of fluid administered and preexisting diseases

SC
Background

The first case reports and case series involving POVL started appearing in the literature in the

U
late 1940s and early 1950s and were related to central retinal artery occlusion (CRAO).[4-6] These case
AN
reports described patients with postoperative unilateral loss of vision which the authors attributed to

direct pressure on the eye, hypotension or a combination of the two. These reports were followed, in
M

1954, by a case series of 8 patients over a 12 year period with unilateral postoperative blindness. All 8
D

patients had neurosurgical procedures which involved a suboccipital or posterior cervical approach in

the sitting or prone positon using the horseshoe headrest.[7] The authors concluded that these cases
TE

were caused by direct pressure on the eye from the headrest which resulted in retinal ischemia. They
EP

were able to replicate these findings in monkeys in which they applied pressure to the globe, lowered

the blood pressure and decreased the circulating blood volume under general anesthesia. After these
C

findings, one of the authors of the study changed the configuration of the horseshoe headrest at their
AC

institution that was routinely used such that there was more space for the face to rest and less chance

of pressure on the eye. With this change to the headrest, the authors reported that in the 2 years

following there were no ocular complications.[7]

CRAO from pressure on the globe decreased over the next several decades as anesthesiologists

became very diligent about checking the eyes during prone cases, mask anesthesia was phased out in
 ACCEPTED MANUSCRIPT

favor of endotracheal intubation, and surgeons gradually shifted away from using the horseshoe

headrest to the Mayfield holder with pins for posterior cervical and suboccipital procedures. Although

CRAO still occurs occasionally with prone procedures, most cases of POVL in the last two decades have

been associated with ischemic optic neuropathy (ION).

PT
In the mid to late 1990s, increasing reports and publications regarding POVL associated with

RI
spine surgery started surfacing with two separate groups publishing on this topic in June of 1997.[3, 8]

The coincident publication of these two articles in the same month and year is indicative of the fact that

SC
the incidence of POVL had suddenly increased. These new reports were in the setting of a change in

spine surgical practice with increased use of complex instrumented fusion for the same diseases,

U
particularly in middle-aged to elderly adults.[9-11] The drive for this change was undoubtedly related to
AN
an aging population with a demand for a more active lifestyle and advances in other perioperative care

areas such as anesthesia and intensive care that made these major operations safer in this population.
M

This change in practice resulted in more cases with longer operative times and higher blood loss in the
D

prone position. Stevens and colleagues retrospectively reviewed 3450 spinal operations and identified

an incidence of POVL of 0.2% with at least four out of these seven cases being diagnosed with ION and
TE

the other three cases with central retinal vein occlusion, air embolism, and an occipital infarction.[3]
EP

These procedures were associated with major spinal fusions, prolonged operative times, large blood loss

and fluid resuscitation and variable degrees of hypotension. The other study published in the same
C

journal in the same month in 1997 was the first case control study on POVL associated with spine
AC

surgery. This case-control study from multiple institutions and case reports revealed that different types

of POVL occurred in these 37 patients including ION (59%), CRAO (24%) and cortical ischemia (8%) and

were also associated with prolonged operative times in the prone position, large estimated blood loss

with large fluid resuscitation and variable degrees of hypotension.[8] The only statistically significant

differences between the affected patients and control patients were the operative duration and the
 ACCEPTED MANUSCRIPT

estimated blood loss. Neither the lowest systolic blood pressure nor the lowest hematocrit differed

between the cases and control groups.

As these studies were published, it became clear to surgeons and anesthesiologists that there

were different types and causes of POVL other than CRAO with globe compression. Williams et al

PT
provided one of the first reviews of POVL after all types of procedures in 1995.[12] The most commonly

RI
diagnosed causes of POVL associated with spine surgery are ION, which is further categorized into

anterior (AION) and posterior (PION), CRAO and cortical blindness. These different ophthalmologic

SC
injuries have different signs and symptoms and are associated with different surgical procedures (Table

1).

U
AN
Ischemic Optic Neuropathy

Ischemic optic neuropathy (ION) can be divided into two types anterior (AION) and posterior
M

(PION). AION is characterized by optic nerve injury occurring anterior to the lamina cribrosa, a piece of
D

sieve-like connective tissue through which the optic nerve and central retinal vessels pass on their way

to the globe. PION is characterized by optic nerve injury posterior to the lamina cribrosa.[13]
TE

AION is most commonly associated with cardiac bypass procedures and major vascular procedures,
EP

prone spine operations, abdominal compartment syndrome, head and neck operations and

miscellaneous types of operations. It can be further divided into arteritic and nonartertitic. Arteritic
C

AION is usually secondary to temporal arteritis and is a rarity in the perioperative period. AION occurring
AC

in the perioperative period is almost always diagnosed as nonarteritic AION and is also the most

common cause of spontaneous sudden visual loss in patients 50 years and older in the general

population.

Patients may present with unilateral or bilateral AION immediately after surgery or may have

normal vision for several days and then present with a sudden painless visual disturbance which
 ACCEPTED MANUSCRIPT

progresses over a few days. Visual field deficits may present as scotoma, altitudinal field cuts or

complete loss of vision with no light perception. Affected eyes will have a relative afferent pupillary

defect with unilateral or asymmetric disease or absent light reflexes. Fundoscopic examination reveals

an edematous optic disc with blurring of the disc margin. Peripheral splinter or peripapillary flame

PT
shaped hemorrhages may also be present. Fundoscopic examination and symptom onset distinguish

RI
AION from PION initially; however, after a few weeks when swelling and hemorrhage diminishes in

AION, the optic disc becomes pale with both lesions and AION and PION appear the same.[13] At this

SC
time there is no treatment proven to be beneficial and recovery of vision is typically poor. [13]

Histopathologic studies of patients who developed PION from bilateral radical neck operations

U
or massive hemorrhage demonstrated lesions in the posterior optic nerve several mm anterior to the
AN
optic canal.[14-16] This area is particularly vulnerable to hypoperfusion because of the minimal overlap

of blood supply in these watershed areas in comparison to the anterior optic nerve. It is currently
M

unclear if the optic nerve ischemia is caused by edema formation and compression of the small pial
D

vessels, high interstitial pressure causing direct injury or venous hemorrhage and / or infarction.[17]

PION is most commonly associated with prone spine operations, bilateral radical head and neck
TE

procedures, and prolonged procedures with the head-down such as robotic urologic and gynecologic
EP

operations. It usually presents as painless loss of vision upon awakening from anesthesia and does not

typically progress. It is associated with procedures with prolonged periods of increased venous pressure
C

in the head such as prone spine surgery, surgeries in steep Trendelenburg position and bilateral head
AC

and neck procedures. PION after spinal surgery has been reported to occur bilaterally more frequently

than unilaterally.[18] Visual field deficits and pupillary light reflexes in patients with PION have a similar

presentation to patients with perioperative AION. Initial fundoscopic examination is normal, but pallor

of the optic disc develops over weeks to months. The recovery from perioperative PION is poor and no

treatment has demonstrated efficacy.[13]

 ACCEPTED MANUSCRIPT

Despite the lack of evidence-based medicine to support therapeutic interventions for AION or

PION, some ophthalmologic consultants will recommend correcting moderate to severe anemia and

restoring blood pressure to awake baseline values. Elevation of the head is also recommended, if not

medically contraindicated, for treatment of any edema formation of the periorbital area and to improve

PT
venous outflow from the head. Other treatment modalities have been attempted with variable success

RI
in case reports including administration of mannitol, high dose steroids or hyperbaric oxygen therapy.[3,

12, 19-21]

SC
Central Retinal Artery Occlusion

U
Central retinal artery occlusion (CRAO) is most often associated with prone spine operations,
AN
cardiac operations and head and neck procedures where injections are performed around the nose and

eyes. It is caused by decreased blood supply to the entire retina. The most common cause
M

perioperatively is improper head positioning which leads to external pressure on the eye. This in turn
D

leads to increases in intraocular pressure (IOP) which can compromise flow of the central retinal artery.

This mechanism of injury is frequently associated with ipsilateral signs of periocular trauma such as
TE

ophthalmoplegia (paralysis of extraocular muscles), ptosis, corneal abrasions, supraorbital nerve injury
EP

with paresthesias and other signs of trauma to soft tissues around the affected eye. The injury is almost

always unilateral. Other causes can include emboli to retinal circulation or arterial thrombosis due to a
C

hypercoagulable state.[22] CRAO presents as unilateral vision loss which is severe and is usually
AC

reported by patients very soon after they are awaken.[23] Pupillary light reflexes will be absent or

slowed with a relative afferent pupillary defect. Classic funduscopic findings of CRAO include a whitened

ischemic retina with a cherry-red spot at the macula indicative of perfusion via the choriocapillaris

vasculature. The prognosis of CRAO is poor and usually results in permanent visual loss; therefore, the

best treatment is prevention. Anesthesia providers should perform periodic eye checks to ensure that
 ACCEPTED MANUSCRIPT

there is no pressure on the globe and surgeons should avoid headrests such as the horseshoe headrest

which leaves little room for the eyes and makes checking eyes difficult during the procedure. Strategies

to minimize embolic loads which can also cause CRAO are well established in cardiac surgery with use of

air filters on the cardiac bypass pumps[24], improved surgical and perfusionist techniques, and epiaortic

PT
ultrasound scanning to avoid placement of cannula through atherosclerotic plaques[25]; but effective

RI
methods to reduce emboli during spine surgery have not yet been developed.

Treatment of CRAO is generally ineffective and there is a paucity of randomized controlled

SC
studies with positive outcomes.[26] Dilating the retinal arterial blood supply and lowering the

intraocular pressure with five percent carbon dioxide in oxygen, intravenous acetazolamide and anterior

U
chamber paracentesis may improve oxygen delivery to the retina.[23, 27] Selective fibrinolytic therapy
AN
via the ophthalmic artery for spontaneous CRAO was proven ineffective compared to conservative

treatment in the randomized European Assessment Group for Lysis in the Eye (EAGLE) Study and
M

complication rates were considerably higher in the fibrinolytic treatment group.[28, 29] One
D

retrospective study for spontaneous CRAO suggested some improvement with hyperbaric oxygen

therapy treatment within 12 hours of onset of symptoms with a mean increase in visual acuity of three
TE

lines[30], though the number of patients was small and randomized controlled trials have not been
EP

performed yet. Treatment within six hours of onset of symptoms for spontaneous CRAO is thought to

provide the highest chance of significant improvement in visual acuity.

C
AC

Cortical Blindness

Cortical blindness is usually associated with either cerebral hypoperfusion in cases with

prolonged profound hypotension or embolism in cases with high embolic loads such as cardiac bypass

procedures.[31] Both are possible during spine surgery, however cortical blindness is the least common

cause of POVL in adults following spine surgery when compared to AION, PION and CRAO. A study of
 ACCEPTED MANUSCRIPT

patients undergoing lumbar spine surgery using transesophageal echocardiography showed that 80% of

patients had moderate to severe embolic events, primarily during instrumentation with pedicle screw

placement.[32] This finding is very similar to studies with internal fixation of femurs; however, the

incidence of cerebral infarctions is very low in both of these groups. Cortical blindness usually presents

PT
on awakening from anesthesia. Examination shows normal pupillary light reflexes and fundoscopic

RI
exam. Unilateral lesions can produce homonymous hemianopia, whereas bilateral occipital cortex

damage can lead to complete blindness. Computed tomography or magnetic resonance imaging of the

SC
brain will reveal any acute lesions. Recovery from cortical blindness is better than other causes of POVL.

Maintenance of normotension and correction of severe anemia is typically recommended for suspected

cases of cortical blindness.

U
AN
ASA POVL Registry
M

In response to the noticeable increase in POVL cases in association with spine surgery in the late
D

1990s, the University Hospital Consortium risk managers contacted the American Society of

Anesthesiologists (ASA). Because POVL is a rare complication and any one institution will likely have a
TE

limited number of cases, the ASA established the ASA POVL registry in 1999. This database allows
EP

anonymous reporting of cases of POVL. The goal was to obtain sufficient numbers of POVL cases to

provide a valuable analysis of patient and perioperative conditions that might be associated with the risk
C

of this condition. In 2006, the analysis of 93 cases of POVL related to spine surgery was published.[33]
AC

Visual loss was caused by ION in 89% of cases (n = 83) and the remaining 10 cases were caused by CRAO.

Of these 83 cases caused by ION, 56 were given the diagnosis of PION, 19 were given the diagnosis of

AION and 8 were unspecified ION. There were no significant differences with respect to patient, surgical

or anesthetic characteristics between the ION cases and all 83 were combined and compared to the 10

cases of CRAO. Patients with ION had statistically significantly higher blood loss, longer anesthetic
 ACCEPTED MANUSCRIPT

duration, and a higher percentage of patients with bilateral disease in comparison to CRAO patients.

Furthermore, 96% of patients with ION had blood loss of one liter or more or an anesthetic time of 6

hours or more.[33]

In 2012, the ASA’s Postoperative Visual Loss Study Group published a multi-institutional case-

PT
control study using the ASA POVL registry cases to identify risk factors associated with ION after spinal

RI
surgery.[17] In the study, 80 patients with ION after spine surgery were compared to controls

undergoing spinal fusions from 17 academic medical centers that were matched for year of surgery.

SC
Statistically significant and independent risk factors for developing ION were male sex, obesity, Wilson

frame use, duration of anesthetic, estimated blood loss (EBL), and decreased percentage of (non-blood)

U
colloid administered (Table 2). The lowest hematocrit and the percentage of patients with blood
AN
pressure more than 40% below baseline for 30 minutes were not significantly different between groups.

This multivariate model had an area under the curve of 85% after cross-validation and a sensitivity and
M

specificity of approximately 80%.

D

Theoretical Pathophysiology
TE

The protective effect of estrogen on the optic nerve was speculated as a possible reason why
EP

females were less likely to have postoperative ION as the anatomy is similar between men and women.

The other five risk factors were potentially related to increased venous pressure in the head with
C

development of interstitial edema. The prone position increases intra-abdominal and intra-thoracic
AC

pressure causing elevated central venous pressure that causes a decrease in venous return and stroke

volume. Obesity associated with a larger abdominal girth will further increase the intra-abdominal and

central venous pressure. The Wilson frame is designed with the head resting considerably lower than

the heart, further increasing venous pressure in the head and exacerbating interstitial edema. Increased

anesthetic time, particularly after five to six hours, is important because this ischemic injury is thought
 ACCEPTED MANUSCRIPT

to be related to an outflow or venous problem with edema formation which will take more time to

develop than an inflow or arterial obstruction. Higher blood loss was presumed to increase the risk of

ION by increasing fluid shifts, inflammation and capillary leak, and lowering oncotic pressure when

blood is replaced with crystalloid. Replacement with colloid will cause less decrement in oncotic

PT
pressure and theoretically diminish the edema formation. It may explain why anemia alone was not

RI
identified as an independent risk factor in the multivariate analysis. Anemia may only be a surrogate

marker for low oncotic pressure. Higher blood loss is also frequently associated with intermittent

SC
reduction in cardiac output and perfusion. These theories regarding the identified risk factors and their

contribution to this pathophysiology will require further research for validation.

U
AN
American Society of Anesthesiologists Practice Advisory

Based on data from the ASA POVL Registry and the associated case-control study, numerous
M

case reports, case series, and multi-disciplinary expert opinions on POVL, the ASA developed a practice
D

advisory in 2006 for the prevention of visual loss associated with spine surgery.[34] Spine surgeons,

neuro-ophthalmologists and anesthesiologists participated in the development of this advisory. This

TE

work was updated in 2012.[35] Perioperative management of head position, headrest, blood loss, fluid
EP

(crystalloid, colloid and blood products) administration, blood pressure, hemodynamic monitoring and

staging of anticipated prolonged procedures with anticipated high blood loss and periodic assessment of
C

eyes and selection of surgical frame type were all considered in this advisory (Table 3).
AC

Because most of the six independent and significant risk factors for ION may not be modifiable,

and the fact that the multivariate predictive model only detects approximately 80% of patients who will

develop ION, and the high likelihood that there are currently undetectable patient-specific risk factors

such as anatomical or physiological aberrancies of the optic nerve circulation or genetic predisposition

to this injury, the experts on the ASA advisory task force determined that consent for ION in patients
 ACCEPTED MANUSCRIPT

who are anticipated to undergo prolonged spine surgery in the prone position and / or with expected

high blood loss should be considered.

Conclusion

PT
POVL is an uncommon, devastating complication associated with spine surgery, and the most

RI
frequent POVL diagnosis is ION. The relatively low incidence of this complication, the ethical limitations

to conducting randomized studies for this outcome measure, and the lack of a current animal model

SC
limit the level of evidence in studies. Current research and expert opinion indicate that this complication

is primarily associated with procedures that create elevated venous pressure in the head for prolonged

U
periods of time. The largest case-control study to date on ION associated with spine surgery found six
AN
independent and significant risk factors including male sex, obesity, use of the Wilson frame, longer

operative times, greater estimated blood loss and a lower colloid to crystalloid ratio in the non-blood
M

fluid administration. Modification of some of these factors may reduce the risk, but supportive evidence
D

is lacking. Avoiding significant physiologic and hemodynamic perturbations as much as possible in this

setting is advisable given the uncertainty of the pathophysiology with this complication. Because
TE

prevention of this complication cannot be guaranteed, consent for POVL should be strongly considered
EP

for patients expected to undergo prolonged spine operations in the prone position and / or procedures

with expected substantial blood loss.

C
AC
 ACCEPTED MANUSCRIPT

Practice Points

• Consenting patients for POVL who are expected to undergo spine operations in the prone
position for prolonged periods of time, have substantial blood loss, or both have a small,
unpredictable risk of perioperative visual loss (POVL) k should be strongly considered.
• Direct pressure on the globe can cause POVL by central retinal artery occlusion (CRAO) and this

PT
risk can be decreased:
o by avoiding headrests such as the horseshoe headrest that do not allow adequate
assessment of eyes during the procedure and have a narrow margin of error;

RI
o by performing periodic eye checks throughout the operation if possible.
• Ischemic optic neuropathy (ION) is the most common cause of POVL associated with prone
spine surgery in adult patients.

SC
• Male sex, obesity, use of the Wilson frame, longer duration, larger blood loss and a lower
percentage of colloid in the non-blood fluid administration significantly and independently
increased the risk of ION associated with spine surgery in the prone position.
•

U
Untested Practice Points that may decrease the risk of ION in high-risk procedures include:
o Minimizing the venous pressure and congestion in the head by:
AN
• Keeping the head neutral and
• at or above the heart level;
• Avoiding use of the Wilson frame if not surgically contraindicated;
M

o Using colloid along with crystalloid for non-blood fluid resuscitation.

o Minimize bleeding by:

Avoiding coagulopathy;
D

Modifying surgical technique

o Decreasing the duration in the prone position by
TE

Staging anticipated very prolonged procedures;

Using well trained surgeons for very prolonged procedures.
o The authors’ practice is to avoid significant physiologic and hemodynamic perturbations
EP

as much as possible, and to continually monitor the blood pressure.

o Resume normotension and normovolemia after turning the patient supine unless
medically contraindicated.
C

Research agenda
AC

• We need an improved mechanism for reporting POVL after all procedures so that the incidence
and trends over time can be better assessed.
• We need more research into development of an intraoperative monitor that can be utilized
under anesthesia in the prone position to assess optic nerve function.
• We need to develop an animal model of ION to better study contributory factors,
pathophysiology, and therapeutic interventions.
• We need to develop a national network for testing therapeutic interventions in the acute stage
of POVL from all causes.
 ACCEPTED MANUSCRIPT

Conflict of interest statement

None.

PT
RI
U SC
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT

References

1. 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. Anesthesia and
Analgesia. 2009;109(5):1534-45.
2. Patil CG, Lad EM, Lad SP et al. Visual loss after spine surgery: a population-based study. Spine.
2008;33(13):1491-6.

PT
3. Stevens WR, Glazer PA, Kelley SD et al. Ophthalmic complications after spinal surgery. Spine.
1997;22(12):1319-24.
4. Slocum HC, O'Neal KC, Allen CR. Neurovascular complications from malposition on the operating

RI
table. Surgery, Gynecology & Obstetrics. 1948;86(6):729-34.
5. Givner I, Jaffe N. Occlusion of the central retinal artery following anesthesia. Archives of
Ophthalmology. 1950;43(2):197-201.
6. Gillan JG. Two cases of unilateral blindness following anaesthesia with vascular hypotension.

SC
Canadian Medical Association Journal. 1953;69(3):294-6.
7. Hollenhorst RW, Svien HJ, Benoit CF. Unilateral blindness occurring during anesthesia for
neurosurgical operations. AMA Archives of Ophthalmology. 1954;52(6):819-30.

U
8. Myers MA, Hamilton SR, Bogosian AJ et al. Visual loss as a complication of spine surgery. A
review of 37 cases. Spine. 1997;22(12):1325-9.
AN
9. Deyo RA, Mirza SK, Martin BI et al. Trends, major medical complications, and charges associated
with surgery for lumbar spinal stenosis in older adults. JAMA. 2010;303(13):1259-65.
10. Deyo RA, Gray DT, Kreuter W et al. United States trends in lumbar fusion surgery for
degenerative conditions. Spine. 2005;30(12):1441-5; discussion 6-7.
M

11. Deyo RA, Nachemson A, Mirza SK. Spinal-fusion surgery - the case for restraint. The New
England Journal of Medicine. 2004;350(7):722-6.
12. Williams EL, Hart WM, Jr., Tempelhoff R. Postoperative ischemic optic neuropathy. Anesthesia
D

and Analgesia. 1995;80(5):1018-29.

13. Lee LA, Newman NJ, Wagner TA et al. Postoperative ischemic optic neuropathy. Spine.
2010;35(9 Suppl):S105-16.
TE

14. Johnson MW, Kincaid MC, Trobe JD. Bilateral retrobulbar optic nerve infarctions after blood loss
and hypotension. A clinicopathologic case study. Ophthalmology. 1987;94(12):1577-84.
15. Nawa Y, Jaques JD, Miller NR et al. Bilateral posterior optic neuropathy after bilateral radical
EP

neck dissection and hypotension. Graefe's Archive for Clinical and Experimental Ophthalmology.
1992;230(4):301-8.
16. Marks SC, Jaques DA, Hirata RM, Saunders JR, Jr. Blindness following bilateral radical neck
dissection. Head & Neck. 1990;12(4):342-5.
C

17. Postoperative Visual Loss Study G. Risk factors associated with ischemic optic neuropathy after
spinal fusion surgery. Anesthesiology. 2012;116(1):15-24.
AC

18. Sadda SR, Nee M, Miller NR et al. Clinical spectrum of posterior ischemic optic neuropathy.
American Journal of Ophthalmology. 2001;132(5):743-50.
19. Ho VT, Newman NJ, Song S et al. Ischemic optic neuropathy following spine surgery. Journal of
Neurosurgical Anesthesiology. 2005;17(1):38-44.
20. Bojic L, Ivanisevic M, Gosovic G. Hyperbaric oxygen therapy in two patients with non-arteritic
anterior optic neuropathy who did not respond to prednisone. Journal of the Undersea and
Hyperbaric Medical Society, Inc. 2002;29(2):86-92.
21. Choudhari NS, George R, Kankaria V, Sunil GT. Anterior ischemic optic neuropathy precipitated
by acute primary angle closure. Indian Journal of Ophthalmology. 2010;58(5):437-40.
 ACCEPTED MANUSCRIPT

22. Newman NJ. Perioperative visual loss after nonocular surgeries. American Journal of
Ophthalmology. 2008;145(4):604-10.
23. Roth S. Perioperative visual loss: what do we know, what can we do? British Journal of
Anaesthesia. 2009;103 Suppl 1:i31-40.
24. Gerriets T, Schwarz N, Sammer G et al. Protecting the brain from gaseous and solid micro-emboli
during coronary artery bypass grafting: a randomized controlled trial. European Heart Journal.
2010;31(3):360-8.

PT
25. Mitchell SJ, Merry AF. Perspective on Cerebral Microemboli in Cardiac Surgery: Significant
Problem or Much Ado About Nothing? The Journal of Extra-Corporeal technology.
2015;47(1):10-5.

RI
26. Fraser SG, Adams W. Interventions for acute non-arteritic central retinal artery occlusion. The
Cochrane Database of Systematic Reviews. 2009(1):CD001989.
27. Wray SH. The management of acute visual failure. Journal of Neurology, Neurosurgery, and
Psychiatry. 1993;56(3):234-40.

SC
28. Schumacher M, Schmidt D, Jurklies B et al. Central retinal artery occlusion: local intra-arterial
fibrinolysis versus conservative treatment, a multicenter randomized trial. Ophthalmology.
2010;117(7):1367-75 e1.

U
29. Pielen A, Pantenburg S, Schmoor C et al. Predictors of prognosis and treatment outcome in
central retinal artery occlusion: local intra-arterial fibrinolysis vs. conservative treatment.
AN
Neuroradiology. 2015;57(10):1055-62.
30. Menzel-Severing J, Siekmann U, Weinberger A et al. Early hyperbaric oxygen treatment for
nonarteritic central retinal artery obstruction. American Journal of Ophthalmology.
2012;153(3):454-9 e2.
M

31. Goni V, Tripathy SK, Goyal T et al. Cortical blindness following spinal surgery: very rare cause of
perioperative vision loss. Asian Spine Journal. 2012;6(4):287-90.
32. Takahashi S, Kitagawa H, Ishii T. Intraoperative pulmonary embolism during spinal
D

instrumentation surgery. A prospective study using transoesophageal echocardiography. The

Journal of Bone and Joint Surgery British Volume. 2003;85(1):90-4.
33. Lee LA, Roth S, Posner KL et al. The American Society of Anesthesiologists Postoperative Visual
TE

Loss Registry: analysis of 93 spine surgery cases with postoperative visual loss. Anesthesiology.
2006;105(4):652-9; quiz 867-8.
34. American Society of Anesthesiologists Task Force on Perioperative B. Practice advisory for
EP

perioperative visual loss associated with spine surgery: a report by the American Society of
Anesthesiologists Task Force on Perioperative Blindness. Anesthesiology. 2006;104(6):1319-28.
35. American Society of Anesthesiologists Task Force on Perioperative Visual L. Practice advisory for
perioperative visual loss associated with spine surgery: an updated report by the American
C

Society of Anesthesiologists Task Force on Perioperative Visual Loss. Anesthesiology.

2012;116(2):274-85.
AC

Table 1: Most Common Causes of Perioperative Visual Loss in Spine Surgery*
POVL Injury Associated
Operations / Conditions
Anterior ION Cardiac bypass, major
vascular, prone spine, head
and neck, abdominal
compartment syndrome.
Posterior ION Prone spine, cardiac, head
and neck, prolonged robotic
cases in the head-down
position.
Central Retinal Prone spine, cardiac bypass,
Artery Occlusion head and neck, high embolic
loads, horseshoe headrest.
Cortical Blindness Cardiac bypass, other
procedures with high embolic
loads (e.g., angiography),
prolonged profound
hypotension / hypoperfusion.
*For any concern of perioperative visual loss, an urgent ophthalmology consultation should be obtained.j
POVL, perioperative visual loss; ION, ischemic optic neuropathy; CRAO, central retinal artery occlusion; RAPD, relative afferent pupillary defect
ACCEPTED MANUSCRIPT
Clinical Presentation Treatment / Prognosis
May have a few days of normal vision before symptoms. No proven therapy.
Painless, progressive loss of vision over several days.
Unilateral or bilateral. Inconsistent results with: normalization of blood
Relative afferent pupillary defect (RAPD) with unilateral or pressure and correction of moderate to severe anemia,
PT
asymmetric disease or absent reflex. high dose steroids, hyperbaric oxygen.
Scotoma, altitudinal field cuts or no light perception.
Early Exam: edematous optic disc, peripheral splinter or Poor prognosis – no to mild recovery of vision.
RI
peripapillary flame-shaped hemorrhages, attenuated vessels.
Late Exam: optic nerve pallor, edema resolved, vessels normal.
Vision loss is not progressive – typically immediately after No proven therapy.
SC
awakening.
Painless. Inconsistent results with: normalization of blood
Usually bilateral. pressure and correction of moderate to severe anemia,
U
Relative afferent pupillary defect (RAPD) with unilateral or high dose steroids, hyperbaric oxygen.
asymmetric disease or absent reflex.
AN
Scotoma, altitudinal field cuts or no light perception. Poor prognosis – no to mild recovery of vision.
Early Exam: normal fundus
Late Exam: optic nerve pallor
Vision loss on awakening from anesthesia. No proven therapy.
M
Unilateral.
Sluggish (RAPD) or absent pupillary light reflex For spontaneous CRAO: inhaled 5% carbon dioxide in
Early Exam: whitened ischemic retina with cherry red spot at oxygen, intravenous acetazolamide and anterior
D
macula. chamber paracentesis.
Late Exam: may have optic nerve pallor, reperfused retina.
TE
Signs of globe compression include ipsilateral periorbital trauma (2 randomized controlled trials showed no benefit and
such as ophthalmoplegia, corneal abrasions, supraorbital increased harm with selective intra-arterial
paresthesias, ptosis, ecchymoses. thrombolysis.)
EP
Poor prognosis – no to mild recovery of vision.
Visual loss on awakening from anesthesia. Normalize blood pressure, volume status, cardiac output
If unilateral, contralateral homonymous hemianopia. and oxygenation to maximize oxygen delivery.
C
If bilateral, small area of central preserved vision or complete loss
of vision. Prognosis better than ION or CRAO, but recovery is not
AC
Normal pupillary light reflexes. typically to baseline.
Normal fundus.
 ACCEPTED MANUSCRIPT

Table 2: Risk Factors for Ischemic Optic Neuropathy Associated with Spine Surgery

Male sex
Obesity
Use of Wilson frame
Anesthesia duration*
Estimated blood loss

PT
Colloid as percent of nonblood fluids
*Anesthesia Duration is a surrogate for surgical duration.

RI
Multivariate regression analysis in a case-control study comparing 80 cases of ION after spine surgery with 315 controls matched by year of surgery and
undergoing spinal fusion procedures in the prone position identified six significant and independent risk factors for ION. [16]

U SC
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT

Table 3: Major Considerations from the ASA Practice Advisory for Perioperative Visual Loss Associated with Spine Surgery (for High Risk Patients)

1. Consider informing patients who are anticipated to have spine operations in the prone position with prolonged duration and / or substantial blood loss
that they have a small but unpredictable risk of perioperative visual loss.
2. Continually monitor systemic blood pressure.
3. Colloids should be used along with crystalloids for maintaining euvolemia.

PT
4. Hemoglobin or hematocrit values should be monitored periodically during surgery, but no optimal transfusion threshold to prevent perioperative visual
loss has been identified.
5. The decision to use alpha-adrenergic agents should be made on a case-by-case basis.

RI
6. Direct pressure on the eye should be avoided to prevent central retinal artery occlusion.
7. Position the head neutral with the face down and the head level or higher than the heart to minimize venous outflow obstruction.

SC
8. Consider staging very prolonged procedures in high risk patients with careful consideration of the risk:benefit ratio.
9. Assess patient vision when the patient becomes alert.

U
10. An urgent ophthalmologic consultation should be obtained if there is concern for postoperative visual loss.
11. Consider optimizing hemoglobin, hematocrit, hemodynamic status and arterial oxygenation.

AN
12. Consider magnetic resonance imaging to rule out intracranial causes of visual loss.
13. Antiplatelet agents, steroids, or intraocular-lowering agents have not been shown to be effective for treatment of perioperative ION.

M
Adapted from American Society of Anesthesiologists Task Force on Perioperative Visual Loss. Practice advisory for perioperative visual loss associated with spine
surgery: an updated report by the American Society of Anesthesiologists Task Force on Perioperative Visual Loss. Anesthesiology. 2012 Feb;116(2):274-8.

D
TE
C EP
AC