The Relationship Between Optic Nerve Sheath Decompression Failure and Intracranial Pressure in Idiopathic Intracranial Hypertension

Ocular neuromyotonia (ONM) is a rare motility disorder in which paroxysms of tonic extraocular muscle contraction from abnormal ocular motor nerve firing result in episodic diplopia and strabismus. Medical therapy with membrane-stabilizing agents has varied success. A surgical approach to treatment has not yet been described. We report the outcomes of strabismus surgery in patients with ONM.; ; We describe 3 patients with sixth nerve paresis and ONM of the affected lateral rectus muscle who underwent strabismus surgery. All patients had a history of radiation therapy for intracranial tumors. Ophthalmologic and orthoptic examinations were performed with appropriate medical and neuroradiologic evaluation. Preoperative and postoperative data are presented and analyzed.; ; Two patients were noted to have ONM after their first strabismus surgery for a sixth nerve palsy. Patients 1 and 2 had 3 surgeries, whereas Patient 3 had 1 operation. Extraocular muscles operated on included the medial rectus and lateral rectus. Preoperative primary gaze baseline esotropia ranged from 35 to 75 prism diopters (Δ). All patients achieved improvement in ocular alignment and motility. Postoperative primary gaze deviations ranged from orthotropia to 20Δ of esotropia. Abduction deficits were unchanged or improved. The follow-up period ranged from 15 months to 2 years.; ; Patients with ONM of a paretic rectus muscle can achieve binocular fusion with strabismus surgery. ONM may manifest postoperatively in patients with a sixth nerve palsy and a contractured medial rectus who, preoperatively, were not noted to have ONM.

Original Contribution The Relationship Between Optic Nerve Sheath Decompression Failure and Intracranial Pressure in Idiopathic Intracranial Hypertension Mark E. Robinson, MD, MPH, Annie Moreau, MD, Ryan O'Meilia, BS, John Pagteilan, BS, Kai Ding, PhD, Raymond Michael Siatkowski, MD, Bradley K. Farris, MD Background: To our knowledge, there are no studies of patients with idiopathic intracranial hypertension (IIH) that address the relationship between level of intracranial pressure (ICP) and likelihood of progressive visual loss despite uncomplicated optic nerve sheath decompression (ONSD). This study investigated whether patients with IIH undergoing ONSD had a higher risk of surgical failure if opening pressure (OP) on lumbar puncture was $50 cm H2O compared to those with OP ,50 cm H2O. Methods: We conducted a retrospective chart review of consecutive patients with IIH who failed maximal medical therapy and underwent ONSD between January, 1992 and November, 2014, and were followed at least 3 months postoperatively. The main outcome measure was the relationship between OP on lumbar puncture and ONSD failure. We also investigated the relationship of OP with visual acuity, visual fields, age, and gender. Results: During this period, 174 patients met inclusion criteria. Of the 40 patients who had an OP $50 cm H2O, 6 (15%) had progressive visual loss after uncomplicated ONSD, vs 6 (4.5%) of 134 patients with an OP ,50 cm H2O (P = 0.032, Fisher exact test). Patients with worse visual acuity at presentation also had a higher risk of progressive visual loss after ONSD (P , 0.001, Cochran-Armitage trend test), as did men (P = 0.048, Fisher exact test). Conclusions: Patients with IIH and an OP $50 cm H2O had a 3-fold increased risk of failure of ONSD to prevent progressive visual loss, requiring a shunting procedure when compared to those with OP ,50 cm H2O. Visual acuity at presentation and male sex also were associated with progressive visual decline after ONSD. These risk factors merit Department of Ophthalmology (MR, AM, RMS, BKF), Dean McGee Eye Institute, University of Oklahoma Health Sciences Center, Oklahoma City, Oklahoma; University of Oklahoma College of Medicine (RO, JP), Oklahoma City, Oklahoma; and University of Oklahoma Health Sciences Center (KD), College of Public Health, Oklahoma City, Oklahoma. Supported in part by an unrestricted grant from Research to Prevent Blindness, Inc, New York, NY. The authors report no conflicts of interest. Address correspondence to Mark Robinson, MD, MPH, Palmetto Health/University of South Carolina, Department of Ophthalmology, 9 Richland Medical Park Dr, Suite 340, Columbia, SC 29203; E-mail: [e-mail] 246 closer follow-up in the postoperative period when signs of further visual deterioration would indicate an urgent need for neurosurgical shunting. Journal of Neuro-Ophthalmology 2016;36:246-251 doi: 10.1097/WNO.0000000000000370 © 2016 by North American Neuro-Ophthalmology Society T o our knowledge, there are no studies of patients with idiopathic intracranial hypertension (IIH) that address the relationship between level of intracranial pressure (ICP) and likelihood of progressive visual loss despite uncomplicated optic nerve sheath decompression (ONSD). There are 2 case reports that describe this phenomenon, and in each case the ICP was .50 cm H2O, and both ONSD and a shunting procedure were required (1,2). Overall, the failure rate of ONSD in preventing progressive visual loss in IIH is low, (5.6% in a previous study of 578 eyes) (3). Although elevated ICP is the primary underlying mechanism causing optic neuropathy, there may be additional contributing factors, such as accumulation of toxic substances within the optic nerve sheath from abnormal flow of cerebrospinal fluid, or as a result of septae or trabeculations within the subarachnoid space (SAS) of the optic nerve (4,5). The purpose of our study was to evaluate the relationship of ICP with the subsequent development of progressive visual loss after uncomplicated ONSD. Confounding variables of preoperative visual acuity and mean deviation (MD) on automated visual field perimetry also were assessed. Understanding the nature of this relationship is important to identify patients at risk for progressive visual loss and determine their optimal management. METHODS After approval by our institutional review board, we conducted a retrospective chart review of 174 consecutive patients with Robinson et al: J Neuro-Ophthalmol 2016; 36: 246-251 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution medically uncontrolled IIH, undergoing ONSD between January, 1992 and November, 2014. Inclusion criteria were diagnosis of IIH based on modified Dandy criteria, recorded opening pressure on lumbar puncture, and subsequent uncomplicated ONSD. Exclusion criteria were previous ventriculoperitoneal (VP) or lumboperitoneal (LP) shunt procedure, presence of cerebral venous sinus thrombosis, or abnormal CSF composition. The definition of "failure of the ONSD to protect vision" was the concern for visual loss leading to referral for a VP or LP shunt within 3 months of the ONSD. Indications for referral for a shunt procedure were progressive loss of visual acuity, visual field, and/or frequent obscurations of vision. Headache was not a referral criterion in this series. All neuro-ophthalmologic examinations and all ONSDs were performed by 1 of 2 primary surgeons (B.K.F. and R.M.S.) by means of a medial transconjunctival approach (3). The primary outcome measure was surgical failure rates in patients with opening pressures ,50 cm H2O vs those with opening pressures of $50 cm H2O. As part of a secondary analysis, we constructed subcategories of opening pressure (,30 cm H2O, 30-39 cm H2O, 40-49 cm H2O and $50 cm H2O) and investigated their relationship to the failure rate. In addition, the following data were collected: initial visual acuity, MD on visual field testing (automated), sex, and age at time of ONSD. Visual acuity data were categorized into 3 groups as follows: 20/40 or better, 20/50 to 20/200, and 20/200 or worse. Visual field MD data were categorized into 2 groups, better than or equal to 220 dB and worse than 220 dB. Descriptive statistics were reported. We used visual acuity and MD of the worse eye of each patient to assess their association with other variables (IIH was thought to be the cause of decreased vision, and no evidence of underlying confounding pathology such as amblyopia or cataract was noted). A sensitivity analysis was also performed based on visual acuity and MD data from both eyes, accounting for within-patient correlation using the generalized estimating equation (GEE) method. Similar results were obtained. The association between 2 binary variables was examined using Fisher exact test, whereas the association between an ordinal variable (e.g., 4-level opening pressure and 3-level visual acuity) and a binary variable was examined using the Cochran-Armitage trend test. Comparison of age at the time of ONSD between groups was made based on a 2sample t test. The Clopper-Pearson 95% confidence interval (CI) was reported for the proportion of surgical failure among patients whose OP was $50 cm H2O and whose vision was 20/200 or worse in at least 1 eye. A 2-sided P-value of ,0.05 defines statistical significance. All analyses were conducted in SAS software (version 9.3, Cary NC). RESULTS A total of 174 patients met inclusion criteria for this study, 87.4% were women. Mean age at diagnosis was 30.0 years Robinson et al: J Neuro-Ophthalmol 2016; 36: 246-251 with a standard deviation of 10.4 years. There was no significant difference in age between those who progressed (mean age 33.3 years) and those who did not (mean age 29.7 years, P = 0.26). The overall rate of progressive visual loss for women after ONSD was 5.3%, and for men, 18.2% (P = 0.048). A worse preoperative visual acuity and higher opening pressure were directly associated. Patients with opening pressure $50 cm H2O tended to have worse visual acuity than patients with opening pressure ,50 cm H2O (P = 0.028, Fig. 1). A worse MD on automated visual fields and higher opening pressure also were directly associated. Fifty-two percent of patients with opening pressure $50 cm H2O had MD worse than or equal to 220 dB, compared to 16% of patients with opening pressure ,50 cm H2O (P = 0.0004, Fig. 2). Of the 40 patients who had an opening pressure $50 cm H2O, 6 (15%) had progressive visual loss after uncomplicated ONSD, vs 6 (4.5%) of 134 patients with an opening pressure ,50 cm H2O (P = 0.032). Analysis of subcategories of opening pressure showed trends of ONSD failure beginning at OP = 30 cm H2O and progressively worsening (P = 0.011, Fig. 3). Failure of ONSD occurred in 33.3% of patients with presenting visual acuity worse than or equal to 20/200 compared with 2.4% of patients with visual acuity better than or equal to 20/40 (P , 0.001, Fig. 4). Although perimetric MD greater than 20 dB was not statistically significant as a predictor of progressive visual loss (16.7% vs 5.7%, P = 0.196), this analysis was limited because the MD data were missing for 46% of the eyes. In patients whose OP was $50 cm H2O and whose vision was 20/200 or worse in at least one eye, the failure rate of ONSD was 50% (95% CI: 18.7-81.3). Of the 12 patients referred for a neurosurgical shunt, 2 improved (20/20 in both patients), 3 stabilized with poor vision (,20/200), 3 worsened, 3 were lost to follow-up, and 1 had an external drain placed without subsequent shunt. Of the 3 patients who underwent ONSD and VP shunt and continued to lose vision, 1 patient progressed to 20/400 in the right eye and had light perception in the left eye, another progressed to finger counts in the right eye and 20/30 in the left eye, and another had no light perception (NLP) in both eyes. Of note, the patient who progressed to NLP after a shunt had bare light perception in the right eye and NLP in the left eye before the shunt and may have progressed to NLP bilaterally before the shunt procedure. DISCUSSION In our study, failure of ONSD to prevent further vision loss was associated with higher OP, male sex, and decreased visual acuity at presentation. Poor visual acuity at presentation previously has been reported to be a risk factor for a poor visual prognosis (6). Failure of ONSD was not 247 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution FIG. 1. Relationship of opening pressure with initial visual acuity in each eye. OP, opening pressure. associated with MD on visual field testing but this is limited by missing data. Patients in our study with an OP $50 cm H2O had a failure rate 3 times that of patients with an OP ,50 cm (15% vs 4.5%), but the strongest predictor of ONSD failure was initial visual acuity. Patients with visual acuity worse than or equal to 20/200 at presentation had a failure rate almost 14 times that of patients with visual acuity better than or equal to 20/40 (33.3% vs 2.4%). Sex might also have predictive value, since men had a failure rate almost 3 times higher than that of women (18.2% vs 5.3%). The combination of poor vision and high OP was associated with the highest ONSD failure rate-patients with OP $50 cm H2O and visual acuity worse than or equal to 20/200 had a rate of progressive visual loss after ONSD of 50%. We recommend that such patients should be monitored very closely in the first month postoperatively, with a low threshold for referral for a neurosurgical shunt. The difference in failure rates between men and women is consistent with that of a retrospective study of 721 patients, which found that severe visual loss in patients with IIH was 2 times more likely in men than women (7). In that study, Bruce et al addressed potential explanations for this finding, including a difference in symptoms between men and women with IIH. Men experienced significantly less headache than women (55% of men vs 75% of women) and the lack of pain may have led to a delay in seeking FIG. 2. Relationship of opening pressure with initial MD on automated perimetry of each eye. MD, mean deviation; OP, opening pressure. 248 Robinson et al: J Neuro-Ophthalmol 2016; 36: 246-251 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution FIG. 3. ONSD failure rate increases as opening pressure increases. ONSD, optic nerve sheath decompression; OP, opening pressure. medical care. In our study, however, the patients were on a standard follow-up schedule after ONSD, so the difference in headache likely had less of an effect, although perhaps the lack of headache led to decreased compliance with acetazolamide. The etiology of progressive visual loss after uncomplicated ONSD remains uncertain because papilledema resolved usually, whereas visual function worsened-of the 12 ONSD failures in our study, papilledema completely resolved in 6, improved in 4, and persisted in 1. Although cisternography or T2 magnetic resonance imaging (MRI) of the orbits was not performed to assess the patency of the surgically created window in the nerve sheath, the resolution of papilledema seemed to indicate that the pressure surrounding the nerve head had been relieved, suggesting ongoing damage to the nerve occurred elsewhere or by a different mechanism. In one patient whose papilledema did not improve after ONSD, the presumed mechanism was closure of the surgically created nerve sheath window. A possible etiology of progressive visual loss after ONSD would be nonorganic visual loss, but there were no indications of this on examination and the patients' optic nerve pallor was generally consistent with their level of visual function. Another possible mechanism of visual loss after ONSD could be a complication of the surgical procedures. This seemed unlikely given the temporal profile of the ensuring visual loss; the vision did not decline immediately after surgery, but rather during the ensuing weeks. FIG. 4. ONSD failure rate increases as initial visual acuity decreases. ONSD, optic nerve sheath decompression; VA, visual acuity. Robinson et al: J Neuro-Ophthalmol 2016; 36: 246-251 249 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution Mauriello et al (8) reviewed the literature and noted that many failures of ONSD occurred within 4 weeks postoperatively, rather than immediately after surgery. Our data generally support this time course. Also, in their and our cases, the vision improved immediately after ONSD, but declined days to weeks later. Another etiology for progressive visual loss after ONSD might have been that although the intraorbital optic nerve had been decompressed, CSF pressure remained high surrounding the intracanalicular or intracranial portion of the nerve. Killer and Subramanian (5) concluded that there is impaired flow from the intracranial SAS to the intraorbital SAS of the optic nerve. This impairment may be bidirectional; some patients with a functioning shunt with normal opening pressure on LP may still have papilledema, demonstrating that the CSF pressure remains high in the SAS of the intraorbital optic nerve despite being lowered in the brain and spinal cord. In addition, some patients undergoing ONSD with resolution of papilledema (demonstrating a lower CSF pressure around the optic nerves) still have a high OP on LP and progressive visual loss. Others have reported cases in which there is progressive visual loss after ONSD that improves after a neurosurgical shunt (9) and progressive visual loss after a shunt that improves after ONSD (10). These reports suggest that, in some patients, one procedure is not better than the other, but that there may be separate CSF compartments that require decompression. Further evidence that there is no free communication between the SAS of the optic nerve and the brain include a concentration gradient of beta trace protein between the brain and intraorbital optic nerve (11), a lack of flow from the brain to the optic nerves of intrathecal contrast on CT cisternography (5), histologic findings of septae and trabeculations within the optic nerve sheaths (5), and cases of asymmetric papilledema where lower-grade papilledema always occurred ipsilateral to the narrower optic canal (12). Killer and Subramanian (13) have provided a detailed summary of the evidence that CSF is compartmentalized between the intraorbital SAS and intracranial SAS. This lack of communication between the SAS of the intraorbital optic nerve and that of the brain, and the occurrence of progressive visual loss after ONSD or after neurosurgical shunt procedure alone, suggests that a high CSF pressure in one of these "compartments" was not decompressed, which may be the cause of continued damage to the optic nerves. In the setting of progressive visual loss after ONSD, this proposed compartmentalization of CSF spaces would, as others have concluded (8,14), argue against repeating an ONSD and instead favor a neurosurgical shunt, in an attempt to decompress CSF pressure in the intracranial space. One limitation of our study is that we had information on only 12 patients with progressive visual loss postoperatively, 3 of whom were lost to follow-up, and 1 patient did not receive a shunt because ICP was not high as measured by an external drain. However, we provide 250 evidence that a neurosurgical shunt can halt progressive visual loss after ONSD, as 5 out of 8 remaining patients stabilized after a shunt procedure. Another limitation of our study was that almost half our patients (46%) lacked perimetric MD data. Also, our retrospective study did not standardize the measurement of opening pressures. Lumbar punctures were, in general, performed in the standard lateral decubitus position, but this was not specifically verified during the chart review nor detailed in every report. Additionally, our patients with progressive visual loss after ONSD were considered at high risk for blindness and therefore were referred urgently to neurosurgery without waiting to schedule a lumbar puncture. In some cases, the neurosurgical service obtained an opening pressure or placed an external drain to measure the ICP before a neurosurgical shunt, but we did not collect these data. In summary, we found that in patients with IIH, an opening CSF pressure $50 cm H2O and/or initial visual acuity worse than or equal to 20/200 were risk factors for continued visual loss after an uncomplicated ONSD. We recommend that such patients be placed on a more rigorous postoperative surveillance schedule for prompt consideration for a neurosurgical shunting procedure to prevent further deterioration of vision. STATEMENT OF AUTHORSHIP Category 1: a. Conception and design: M. E. Robinson, A. Moreau, K. Ding, R. M. Siatkowski, B. K. Farris; b. Acquisition of data: M. E. Robinson, A. Moreau, R. O'Meilia, J. Pagteilan; c. Analysis and interpretation of data: M. E. Robinson, A. Moreau, R. O'Meilia, J. Pagteilan, K. Ding, R. M. Siatkowski, B. K. Farris. Category 2: a. Drafting the article: M. E. Robinson, R. O'Meilia, J. Pagteilan; b. Revising it for intellectual content: A. Moreau, K. Ding, R. M. Siatkowski, B. K. Farris. Category 3: a. Final approval of the completed article: M. E. Robinson, A. Moreau, R. O'Meilia, J. Pagteilan, K. Ding, R. M. Siatkowski, B. K. Farris. REFERENCES 1. Wilkes BN, Siatkowski RM. Progressive optic neuropathy in idiopathic intracranial hypertension after optic nerve sheath fenestration. J Neuroophthalmol. 2009;29:281-283. 2. Linden JA, Siatkowski RM. Progressive postoperative visual loss in idiopathic intracranial hypertension with extremely elevated ICP. J Neuroophthalmol. 2010;30:386-387. 3. Moreau A, Lao K, Farris B. Optic nerve sheath decompression: a surgical technique with minimal operative complications. J Neuroophthalmol. 2014;34:34-38. 4. Killer HE, Jaggi G, Miller N. Progressive optic neuropathy in idiopathic intracranial hypertension after optic nerve sheath fenestration. J Neuroophthalmol. 2010;30:205-206. 5. Killer HE, Jaggi GP, Flammer J, Miller NR, Huber AR, Mironov A. Cerebrospinal fluid dynamics between intracranial and the subarachnoid space of the optic nerve. Is it always bidirectional? Brain. 2007;130:514-520. 6. Corbett JJ, Savino PJ, Thompson S, Kansu T, Schatz NJ, Orr LS, Hopson D. Visual loss in pseudotumor cerebri. Followup of 57 patients from five to 41 years and a profile of 14 patients with permanent severe visual loss. Arch Neurol. 1982;30:461-474. Robinson et al: J Neuro-Ophthalmol 2016; 36: 246-251 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution 7. Bruce BB, Kedar S, Van Stavern GP, Monaghan D, Acierno MD, Braswell RA, Preechawat P, Corbett JJ, Newman NJ, Biousse V. Idiopathic intracranial hypertension in men. Neurology. 2009;72:304-309. 8. Mauriello JA Jr, Shaderowfsky P, Gizzi M, Frohman L. Management of visual loss after optic nerve sheath decompression in patients with pseudotumor cerebri. Ophthalmology. 1995;102:441-445. 9. Spoor TC, McHenry JG. Long-term effectiveness of optic nerve sheath decompression for pseudotumor cerebri. Arch Ophthalmol. 1993;111:632-635. 10. Kelman SE, Sergott RC, Cioffi A, Savino PJ, Bosley TM, Elman MJ. Modified optic nerve sheath decompression in Robinson et al: J Neuro-Ophthalmol 2016; 36: 246-251 11. 12. 13. 14. patients with functioning lumboperitoneal shunts and progressive visual loss. Ophthalmology. 1991;98:1449- 1453. Killer HE, Jagger GP, Flammer J, Miller NR, Huber AR. The optic nerve: a new window into cerebrospinal fluid composition? Brain. 2006;129:1027-1030. Bidot S, Bruce BB, Saindane AM, Newman NJ, Biousse V. Asymmetric papilledema in idiopathic intracranial hypertension. J Neuroophthalmol. 2015;35:31-36. Killer HE, Subramanian PS. Compartmentalized cerebrospinal fluid. Int Ophthalmol Clin. 2014;54:95-102. Brazis P. Surgery for intracranial hypertension. J Neuroophthalmol. 2009;29:271-274. 251 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited.