Visual recovery after unilateral optic nerve sheath fenestration for pseudotumor cerebri syndrome

ABSTRACT Purpose To evaluate the effect of optic nerve sheath fenestration (ONSF) on the rate of visual function improvement in patients with pseudotumor cerebri syndrome (PTCS). Methods Retrospective chart review of patients with PTCS who underwent ONSF between 1998 and 2017. Visual function was evaluated by evaluating visual field (VF), mean deviation (MD), retinal nerve fiber layer (RNFL) thickness, papilledema grade, and visual acuity (VA) prior to and after ONSF. Results Seventeen female patients aged 17 to 36 years underwent unilateral ONSF. Follow-up averaged 40.1 months. VF MD improved steadily in both eyes up to 12 months. Average RNFL thickness improved in the operated eye from 347 ± 166 mm to 92 ± 27 mm (p < .001) and the non-operated eyes from 306 ± 165 mm to 109 ± 46 mm (p < .001). The grade of papilledema improved in the operated eye from 3.3 ± 1.3 to 0.3 ± 0.7 and the non-operated eye from 3.0 ± 1.6 to 0.18 ± 0.4. There was an exponential rate of improvement in papilledema and RNFL thickness, with the greatest improvement occurring within the first 30 days. Average visual acuity remained intact in both eyes before and after surgery. Conclusions ONSF in appropriately selected patients leads to rapid improvement in papilledema and a steady recovery in VF.


Introduction
Pseudotumor cerebri syndrome (PTCS) is a disorder characterized by both primary (e.g. idiopathic intracranial hypertension) and secondary (e.g. medication induced) causes of elevated intracranial pressure (ICP). 1 It is most common in young, obese women. 2 Common symptoms include headaches, transient visual obscurations, pulsatile tinnitus, and rarely diplopia. Vision loss is a notable complication, with blindness in up to 10% of patients and some degree of visual field loss in up to 90% of patients. 3,4 The goal of treatment is to decrease the ICP to improve symptoms and prevent further vision loss. Most patients can be managed conservatively by a combination of weight loss and medications such as acetazolamide. 5 Surgical intervention is indicated in cases of worsening vision and symptoms refractory to medical management; options include cerebrospinal fluid (CSF) diversion procedures such as ventriculoperitoneal (VP) or lumbar-peritoneal (LP) shunts, venous sinus stenting (VSS), and optic nerve sheath fenestration (ONSF). 6 ONSF has been shown to be a safe and effective procedure in averting further visual decline, often used in cases of fulminant increased intracranial hypertension. [7][8][9][10] However, there is a gap in current literature describing the course of visual function recovery following ONSF. In this study, we report the rate of visual function improvement in both the operated and non-operated eyes of patients with PTCS who underwent unilateral OSNF.

Materials and methods
The research followed the tenets of the Declaration of Helsinki and was approved by the institutional review board of the Lahey Hospital (Burlington, MA). Information was gathered and secured in compliance with the Health Insurance Portability and Accountability Act. The requirement for informed consent was waived because of the retrospective nature of the study. The medical records of patients who underwent ONSF for PTCS at a tertiary medical center from 1998 to 2017 were retrospectively reviewed. Patients who met the modified Dandy criteria for idiopathic intracranial hypertension (IIH) and patients with medication induced increased ICP were included. 11 Patients with less than three months of postoperative follow-up were excluded from the study. All patients underwent comprehensive neuro-ophthalmologic examination and functional testing, including optical coherence tomography (OCT) and visual field (VF), both pre-and post-operatively. The decision to perform ONSF was determined by the neuro-ophthalmologist based on progressive visual dysfunction despite maximal medical treatment and/or rapidly worsening vision at time of initial presentation.
The optic nerve of the more severely affected eye was chosen for unilateral ONSF to avoid surgical risk to the better functioning eye. ONSF was performed under general anesthesia by a single surgeon (ST) using a superomedial lid crease approach in all cases. 12 Expression of CSF was observed following initial incision into the dural sheath and a 3 × 2 mm segment of dura and arachnoid was excised between 3 and 4 mm from the globe. Postoperative data were obtained at intervals of 1 week, 1 month, 3 months, 6 months, 12 months, and at longest follow-up and included the results of neuro-ophthalmic examination, VF testing (30-2 SITA standard protocol; Humphrey visual field analyzer, Carl Zeiss Meditec Inc., Dublin, CA), and average retinal nerve fiber layer (RNFL) thickness on OCT testing (CIRRUS, Carl Zeiss Meditec Inc.).
Primary outcomes included VF mean deviation (MD), OCT-RNFL thickness, grade of papilledema, and visual acuity (VA) compared at each follow-up visit. Papilledema was graded clinically based on the Frisen papilledema scale by a neuro-ophthalmologist (GA). 11 Secondary outcomes measures included headaches, transient visual obscurations, and stopping of medical treatment after ONSF. The data were analyzed by paired and independent t-tests, analysis of covariance (ANCOVA), and Pearson product-moment correlation coefficient using IBM SPSS (version 28.0, SPSS, Inc., Chicago, IL). The datasets generated during and/or analyzed during the current study are available from the corresponding author (YD) on reasonable request.

Results
Seventeen patients, all of whom were female, were identified. Age ranged from 17 to 52 years old (average age 26 ± 10 years). All patients underwent unilateral ONSF. Follow-up ranged from 3 to 120 months (average 41 ± 40 months). One patient had a previous shunt surgery requiring multiple revisions. Two patients reported prior minocycline use. All patients were on at least one medical treatment before undergoing ONSF (16 on acetazolamide, 1 on topiramate, and 3 on furosemide).
Sixteen of these patients (94%) had a successful clinical outcome based on a significant improvement in papilledema, improvement of VF, and no evidence of further vision loss. One remaining patient had further acuity loss despite a dramatic improvement in papilledema in both eyes immediately following ONSF. This patient therefore underwent transverse venous stenting 48 days after ONSF with no subsequent recovery of visual function. Excluding this patient, as well as one non-operated, fellow eye from a patient who had underlying optic atrophy at baseline, VA remained stable before and after ONSF. The average logMAR VA remained stable from 0.40 ± 0.67 (Snellen equivalent of 20/50) to 0.18 ± 0.32 (Snellen equivalent of 20/30; p = .179; n = 17) in the operated eye and from 0.20 ± 0.57 (Snellen equivalent of 20/32) to 0.054 ± 0.24 (Snellen equivalent of 20/22; p = .161; n = 16) in the nonoperated eye. The thickness of the RNFL decreased exponentially in both the operated and non-operated eyes for which adequate data were available ( Figure 1). At last follow-up, RNFL thickness improved in the operated eyes from 347 ± 166um to 92 ± 27 um (p < .001; n = 14) and the non-operated eyes from 306 ± 165um to 109 ± 46 um (p < .001; n = 13). Excluding one non-operated eye that had baseline optic atrophy, the mean Frisen papilledema grade improved in both the operated eyes from 3.3 ± 1.3 to 0.3 ± 0.7 (p < .001) and the non-operated eyes from 3.0 ± 1.6 to 0.19 ± 0.4 (p < .001). Similarly, the papilledema grade demonstrated an exponential decline following ONSF in both eyes, with the greatest improvement occurring within one month ( Figure 2). Fourteen patients had adequate VF studies, defined as having a baseline VF and at least two additional VF at subsequent follow-up visits. Among these patients, 71% experienced an improvement in their VF-MD in one or both eyes. For the subset of patients who demonstrated improvement, the MD for both the operated (n = 10) and the non-operated eyes (n = 9) steadily improved post-operatively. However, the non-operated eye achieved significantly greater improvement as percent recovery in the MD (F 1,60 = 12.521, P < .001). The rate of improvement in the VF-MD was similar for both the operated and the non-operated eyes (ANCOVA, F 1,60 = 0.046, P = .830; Figure 3; Supplemental Table S1).
Postoperatively, the majority of patients (69%) reported improvement in headaches and resolution of TVOs (87%) by 1-month post-surgery. Ten patients (62%) were able to stop oral acetazolamide use after ONSF within a range of 0 days to 15 months (Table 1). Within this group of patients, the VF-MD trended toward improvement in both the operated and non-operated eyes (−11.20 to −9.01 and −7.08 to −4.26, respectively). In comparison, the group of patients who continued medical treatment after ONSF trended toward worsening and/or stabilization of VF-MD in both the operated and non-operated eyes (−5.92 to −8.74 and −5.01 to −5.61, respectively). In total, six patients (35%) experienced minor complications after ONSF including lid irritation, ptosis, diplopia, and small hematoma in the immediate post-operative period. All complications were transient in nature and resolved completely within one month after surgery.

Discussion
Most patients with PTCS have a good prognosis with medical management. However, in some cases, preserving visual function requires timely surgical intervention. ONSF is an effective surgical option in patients with worsening visual symptoms who are refractive to medical management without severe optic atrophy. Similar to previous studies, 81% of our patients experienced stabilization or improvement of VA. 7,13 Although many studies have  reported improvement in papilledema after ONSF, the rate of visual function improvement is not well described. 8,9,13 Our study demonstrated rapid improvement of papilledema, both clinically and when measured by OCT of the RNFL, as early as one week after surgery in both the operated and non-operated eyes. By comparison, VF improvement showed a slower, gradual improvement over a 12-month postoperative period. This information is helpful when judging the surgical success of a patient with PTCS, as well as determining if additional treatment is required.
The severity of papilledema often correlates with visual function. However, in cases where there is optic atrophy and pallor, visual function may remain poor despite improvement of papilledema. Patients with features of chronic papilledema and optic atrophy already have limited numbers of functioning axons and are therefore at a higher risk of additional nerve damage from surgical manipulation (e.g. stretch injury). 14 Such nerves are often too delicate to withstand surgery. This is reflected by one such patient in our cases series who initially presented with pallid edema and did not demonstrate any visual function recovery after ONSF. Similar to the results by Alsuhaibani et al., the patients in our study had significant decrease in papilledema grades in both the operated and non-operated eyes following unilateral ONSF. 8 There are reports of both immediate and long-term effects of ONSF on cerebrospinal fluid (CSF) accumulation surrounding the optic nerve. 15,16 Although the mechanism for bilateral and sustained improvement of papilledema is not entirely clear, it is believed that creation of a window in one optic nerve allows for persistent outflow of CSF from the subarachnoid space of both optic nerves. Magnetic resonance imaging studies of the optic nerve after ONSF show cystic structures around the nerve within the immediate post-operative period that later disappear, suggesting CSF filtration as the dominant mechanism during the early post-operative period. 17 In an anatomical study by Ochoa et al. using electron microscopy, a disruption of the node of Ranvier was noted in peripheral nerves of baboons subjected to a compression tourniquet. 18 In this study, prolonged compression led to demyelination beginning at day seven and the extent of pathological changes correlated with the duration and severity of nerve compression. The changes at the level of the Schwann cell junction led to the obliteration of the nodal gap and subsequent conduction block. Furthermore, this study illustrated that re-myelination following compression release occurred over several months and was associated with intra-myelin and periaxonal edema that resulted in localized swelling. Based on these observations, decompression of the optic nerve sheath by means of ONSF may allow for the reformation of nodal gaps and restoration of saltatory conduction. This could be a possible explanation for the delay in visual field improvement despite a rapid and visible reduction in optic nerve swelling.
Optic nerve sheath fenestration has not been traditionally recommended as a treatment for headaches associated with PTCS. The Idiopathic Intracranial Hypertension Treatment Trial (IIHTT) showed no correlation between CSF pressure and the presence or severity of headaches; most patients also continue to have headaches despite normalization of CSF pressure. 19 In our study, 69% of patients reported resolution of their headaches, which is an even greater proportion of affected patients compared with the 41% reported in the review by Kalyvas et al. 20 Many patients are believed to have a mixed headache syndrome and the presence of headaches despite improvement or resolution of papilledema suggests that chronic headaches in patients with PTCS is unlikely to be due to increased ICP alone. In Feldon's meta-analysis of 423 eyes that underwent ONSF, 12% of eyes underwent reoperation. 6 In our study, only one patient required subsequent transverse VSS because of a progressive decline in visual function after having initially improved. No patient underwent a second fenestration procedure. In patients with more severe bilateral disease, bilateral optic nerve sheath fenestrations may be necessary, especially if the surgical technique involves creation of a slit versus window.
Optic nerve sheath fenestration has an overall low complication rate, the majority of which are minor and transient. 7,20,21 However, a small percentage of patients may experience complications, including orbital apex syndrome and traumatic optic neuropathy, that can result in severe loss of vision. 21 23 The reported major complication rate (1.5%) is also lower than that of CSFD (7.6%) and VSS (2.9%). CSFD is fraught with high revision rates, with 15% of patients needing shunt revision within the first 30 days of surgery, as well as high morbidity and mortality rates. 23,24 In addition, progressive VF loss can still occur despite functioning CSF shunts. 7 Although VSS has an overall low complication rate, there have been reported cases of mortality due to severe hemorrhages. 23 Similar to findings by Plotnik & Kosmorsky, all complications in our series were minor and transient with no new vision loss or strabismus. 21 The perioperative cost of ONSF is less than a quarter than that of CSFD and VSS. 20 The utilization rate of ONSF is significantly less than that of CSFD, largely due to the limited availability of surgical expertise required to perform ONSF. 25 As the incidence of IIH has been highly correlated with the rise in obesity rate, the disease burden of PTCS is expected to grow in tandem with the obesity rate, which is projected to reach nearly 50% of the total US population by 2030. 26,27 Therefore, it is important to highlight ONSF as a safe and cost-effective surgical intervention capable of preventing further vision loss in select patients with PTCS.
We recognize the limitations of our retrospective study, which includes a relatively small sample size with variability in follow-up times. Several demographic and outcome  (mo)  1  29  Right  +  -+  -+  -0 days  45  2  20  Left  +  -+  -+  -5 months  11  3  19  Left  +  -+  -+  +  12  4  20  Right  +  -+  -+  -12 months  114  5  17  Left  +  -+  -+  +  98  6  19  Right  +  +  +  -+  +  15  7  23  Left  +  -+  -+  -15 months  21  8  18  Left  +  -+  -+  +  13  9 28 measures, including weight changes and color vision, were also unavailable in this study but may be worthwhile to look at in future ONSF studies. There is also an inherent selection bias in the patients recommended to undergo ONSF, as patients who had chronic papilledema, optic atrophy and signs of optic nerve ischemia (e.g. cotton wool spots, pallid edema) were typically not considered good candidates for the procedure. Our findings may therefore not extend to these patients or others with severe optic nerve injury. Future studies looking at the time from diagnosis of PTCS to surgical intervention may offer further insights into the effect of this procedure on visual outcomes.
In conclusion, our study demonstrates that for patients with PTCS resistant to medical therapy and associated with declining vision, unilateral OSNF can lead to significant improvement in visual function in both eyes. This is supported by the rapid and durable improvement in papilledema and RNFL thickness, as well as slow, but steady recovery in VF that follows these procedures.