Optic Nerve Sheath Fenestration for Progressive Visual Loss in Cerebral Venous Sinus Thrombosis: A Long-Term Retrospective Observational Study

This article is published with digital features, including procedural videos, to facilitate understanding of the article. To view digital features for this article go to https://​doi.​org/​10.​6084/​m9.​figshare.​21723761.

Cerebral venous sinus thrombosis (CVST) is a cerebrovascular disease characterized by occlusion of cerebral veins or sinuses and the subsequent development of intracranial hypertension (IH), owing to impaired cerebrospinal fluid (CSF) absorption [1]. Medical treatments of CVST include anticoagulants, intracranial pressure (ICP)-lowering medications, and repeated lumbar punctures, effectively reducing CVST’s mortality rate [2]. Nevertheless, CVST carries a potential risk of progressive vision loss. Despite receiving maximum medical therapy, 6.7–40% of patients are estimated to suffer from visual impairment and even progress to complete blindness [3‐6], especially those with diagnostic delay or chronic IH, at least partly due to prolonged optic nerve compression [7, 8]. Furthermore, since CVST is more common among young and middle-aged people [9], visual impairment often significantly impacts family and socioeconomic development. Therefore, it is critical for neurologists, ophthalmologists, and neurosurgeons to consider surgery when visual loss progresses, despite maximum medical therapy. Optic nerve sheath fenestration (ONSF), which involves cutting slits or windows in the optic nerve sheath to allow CSF to escape, thus relieving optic nerve compression, has been shown to be effective and safe in protecting visual function of patients with idiopathic intracranial hypertension (IIH) [10‐13]. However, the mechanism of visual loss in CVST including primary venous ischemia and secondary intracranial hypertension is more complicated than that in IIH. It is unclear whether ONSF could help improve visual impairment in CVST. Concerns about surgical risks of intraoperative bleeding from anticoagulation greatly limit the use of ONSF in CVST, as most patients have been receiving long-term anticoagulation therapy since diagnosis, so few sporadic cases with relatively short-term follow-ups describing ONSF utility in CVST have been reported [3, 14, 15]. Therefore, an evaluation of ONSF’s efficacy and safety for CVST based on long-term observations is of paramount important, while anticoagulants are used simultaneously. Our purpose is to evaluate the efficacy and safety of ONSF in resolving CVST-induced visual impairment based on long-term observation. Here, we present our long-term follow-up results of ONSF in a population of CVST-induced progressive visual loss that is larger than previously studied.

A total of 31 eyes from 18 patients who had undergone ONSF were included in the study. Their demographic and clinical characteristics are summarized in Table 1. Of the 18 patients, 9 were women (50%) and 9 were men (50%). Their mean age was 34 (range 8–53) years for women and 44 (range 4–69) years for men. The follow-up durations ranged from 1 week to 8 years, with a mean of 35.6 months. All patients underwent initial postoperative assessments 1 week after ONSF. Subsequently, except for two patients who were lost to follow-up (nos. 1 and 2), 16 patients underwent long-term follow-up for at least 6 months, with a mean of 40 months (range 6–96 months). The median duration of visual loss was 3.5 months (range 0.5–13 months). Other symptoms included headache, tinnitus, and diplopia. The median duration from onset to ONSF was 6.5 months (range 2–36 months).

All patients were treated with ICP-lowering medications such as acetazolamide and mannitol; 16 were treated with anticoagulants, and 2 had discontinued oral anticoagulants for a long time. Of the patients, 7 (38.9%) underwent catheter-directed thrombolysis before ONSF. Causes or risk factors were identified in 12 patients, and 6 had no identifiable cause, despite extensive laboratory examination. The predominant cause of CVST was the patient’s underlying disease, which induced various hypercoagulable states

MRI/MRV imaging revealed that the following vessels were involved, in descending order: the transverse sinus, sigmoid sinus, superior sagittal sinus, torcular herophili, straight sinus, inferior sagittal sinus, internal jugular vein, and cavernous sinus. Moreover, 13 (72.2%) patients had multiple sinus thromboses, and 5 (27.8%) had single sinus thrombosis.

This is the largest observational study involving long-term follow-up to assess the efficacy and safety of ONSF in CVST. ONSF was 80.6% (1 week postoperatively) and 60.7% (after long-term follow-up) effective in stabilizing and/or improving visual function, as well as 100% effective in papilledema resolution. Moreover, it exhibited a favorable safety profile with an extremely low complication rate of 5.6% under perioperative anticoagulation. Our study provides essential knowledge for preventing or reversing progressive visual loss in CVST via ONSF.

Visual loss in CVST is often insidious, manifesting as transient amaurosis, blurred vision, and mild visual impairment in early stages of disease. If prolonged, ischemia or optic nerve compression potentially progresses and manifests as visual-acuity loss and/or visual-field defects. Although the International Study on Cerebral Vein and Dural Sinus Thrombosis revealed that the visual impairment in CVST isn’t severe, its frequency may be underestimated owing to the sole inclusion of severe visual loss (< 20/50) [6] and the use of the confrontation testing rather than the automated perimetry or optometric evaluation [6, 7]. Recently, Liu et al. [3] found that 40% of patients experienced permanent visual-field deficits detected by automated perimetry. They also found that 21.5% of patients demonstrated worsening papilledema on follow-up examinations despite maximum treatment. These patients were at high risk of vision loss with worsening papilledema in chronically elevated ICP [3]. Similarly, another study demonstrated that, at the 1-year follow-up, 51.7% of patients presented with worse ocular presentation than onset, and 22.4% of them presented with severe vision loss due to optic atrophy [5]. In our study, all patients continued to exhibit varying degrees of BCVA and/or VFs impairment, despite adequate anticoagulation and ICP-lowering therapy: 11 of 18 patients (61.1%) presented with a preoperative BCVA worse than 20/800, and 6 of 18 patients (33.3%) presented with “NLP.” Thus, visual functions of patients should be closely monitored, and visual impairment actively prevented and treated.

The pathogenesis of visual loss in CVST involves venous infarction and compression from secondary IH due to impaired CSF absorption [1]. Elevated ICP transmitted to the optic nerve subarachnoid space leads to optic nerve sheath hypertension and impairs axoplasmic transport and blood supply to the optic nerve [3, 18, 19]. In our study, all patients experienced an ICP > 300 mmH₂O, except for one patient (no. 5) with an ICP of 210 mmH₂O.

Treatment options for CVST-induced visual loss in IH range from conservative to surgical. Conservative strategies include ICP-lowering medications (acetazolamide, mannitol, and furosemide) and serial lumbar punctures [2]. Acetazolamide, a carbonic anhydrase inhibitor, potentially reduces ICP by lowering CSF production; nonetheless, its effect is limited [20]. Two diuretics, mannitol and furosemide, can reduce ICP via osmotic dehydration. However, excessive dehydration potentially leads to blood concentration and carries the theoretical risk of thrombosis [1]. Serial lumbar punctures may be necessary when hypertension is persistent; nevertheless, lumbar puncture typically requires temporary cessation of anticoagulants, with an attendant risk of thrombosis [20]. Despite these measures, some patients may continue experiencing progressive visual loss, like in our study [3‐6]. Surgery, including shunting procedures and ONSF, is usually reserved for patients who have failed conservative therapy. No consensus has been reached regarding the optimal surgical treatment, and the choice of surgical intervention is largely dependent on local availability and expertise [21]. A systemic review revealed that shunting is ineffective in preventing death or severe disability in acute CVST [22]. Additionally, the safety of shunting is also a concern, as it can be complicated by stroke, intracranial hemorrhage, catheter malposition, CSF infection, and secondary Chiari malformation [23].

On this premise, ONSF seems to have an advantageous edge. It involves cutting slits or windows in the optic nerve sheath to allow CSF to escape and relieve optic nerve sheath hypertension-induced optic nerve injury [18, 19]. It effectively protects visual function secondary to optic nerve sheath hypertension, such as in patients with IIH and related conditions [10, 23]. Compared to shunting, ONSF has been reported to have advantages of higher rates of vision improvement, fewer complications, faster recovery, lower costs, and shorter operative time [24, 25]. In our study, the operative time was approximately 30–40 min per eye. Furthermore, we found an efficacy of 80.6% in stabilizing and/or improving visual function 1 week postoperatively and 60.7% after long-term follow-up. Since the visual function of CVST is mainly affected by two factors: (1) venous ischemia directly caused by CVST and (2) IH secondary to impaired CSF drainage, we speculated that the difference of efficacy between short- and long-term follow-up may be resulted from continued visual loss due to persistent venous ischemia.

It should be noted that we defined both the improvement and maintenance of BCVA as “Effectiveness”, which is because all the patients included in the present study were experiencing progressive visual loss despite adequate anticoagulation and ICP-lowering therapy before ONSF. Therefore, the maintenance of BCVA after ONSF indicates that ONSF may halt this deteriorating process, and could be considered effective in hindering the progressive visual loss in CVST. Noteworthily, among eight eyes with an “NLP” status before ONSF, four improved to “LP,” “HM,” or “CF” status soon after ONSF (patient NO.7, 15, 16), although another four were still in “NLP” status, potentially attributable to irreversible optic nerve damage. This indicates that ONSF plays a positive role in relieving the pressure around the optic nerve and in resolving severe visual loss in CVST patients.

It was reported that the visual acuity required for visual field examination is over 20/200, and for patients with poor central visual acuity the visual field cannot be accurately and precisely detected [26]. Similarly, in our study, 11 patients failed to undergo preoperative and postoperative VFs examinations due to severe visual impairment and even blindness. Therefore, we only presented the VFs as supplementary results to visual acuity. An interesting phenomenon is that, compared with BCVA, the visual field did not show any significant changes in the follow-ups. Our finding is consistent with the results of study by Liu [3]. They found that only 7.7% patients had decreased visual acuity, while 40% had permanent visual field deficits, which may mean that visual field defects are more serious in CVST than visual acuity loss. There are two possible reasons for this phenomenon: (1) there is selection bias because only patients with good vision have VFs, and for patients with poor central vision, the VFs cannot be detected, and (2) the peripheral visual field is more susceptible to compressive ischemia than the central visual acuity. This leads to the peripheral visual field being first damaged from the compression by IH, while the central visual acuity might remain intact [19].

Many studies have verified ONSF safety in IIH, especially after 2000, with surgical procedures development and new instruments application. Compared with the relatively high complication rates of 44.7% and 45.3% reported, respectively, in 1993 [27] and 2000 [10], the complication rate decreased significantly, ranging from 2.6 to 23.0% [28, 29]. Most complications thereafter have been transient and benign without severe sequelae, as summarized in Table 3 [10‐13, 27‐29]. Moreover, ONSF appears to have a lower major complication rate of 2% compared with shunting, with a major complication rate of 8%, according to a meta-analysis of medically refractory IIH [25].

Although ONSF is effective and safe, it targets the compression of optic nerve secondary to impaired CSF drainage and IH. We emphasize that the anticoagulant is still a very basic and important treatment for venous ischemia directly caused by CVST, whereas ONSF can be considered an adjunctive treatment for progressive vision loss of CVST patients in combination with anticoagulation and ICP-lowering therapy. However, concerns regarding the use of perioperative anticoagulants have limited the application of ONSF in CVST. Whether to continue anticoagulation treatment perioperatively is a common clinical dilemma, and surgeons must balance vision-threatening hemorrhage risk against life-threatening thrombo-embolic events. This study used perioperative anticoagulation and bridging strategies to investigate ONSF safety under perioperative anticoagulation for CVST. The results showed a low complication rate of 5.6% without hemorrhagic events. This indicates that ONSF for CVST under perioperative anticoagulation is safe and practicable. Likewise, our perioperative anticoagulation and bridging strategies might be applied to lumber puncture, which usually accompanies a similar anticoagulation dilemma [30]. Notwithstanding, further studies are warranted.

Although preoperative questionnaire was not administered, most patients were in extreme frustration, anxiety, and desperation upon admission because of fulminant progressive visual loss. Visual function cannot fully reflect psychological distress, social isolation, fear, or negative effects on daily life. The ADL and VFQ-25 evaluations were added at the final follow-up to multidimensionally assess the impact of visual problems on patients. ADL results revealed that most patients maintained basic self-care ability in life, while VFQ-25 results indicated that their QoL was significantly affected by visual loss, especially regarding dependency, driving, distance activity, social function, and mental health.

Notably, these patients had visited three to four hospitals before receiving ONSF, and five (27.8%) had a vision-loss duration of over 6 months. Prolonged IH and delayed ONSF treatment appeared to negatively impact visual function [24], we recommend that clinicians might consider ONSF earlier for patients with progressive visual loss. In future investigations, a prospective, randomized controlled trial comparing surgical treatment options (including ONSF and shunting) for CVST-induced visual loss is required. The biggest strength of our study is that we are the first to report the long-term follow-up results of ONSF in a larger population of CVST-induced progressive visual loss than previously studied, while anticoagulants were used simultaneously. However, our study has a few limitations. First, it was subject to selection bias during data collection. The 18 patients we retrieved were referred to our ophthalmology department from neurology departments of various hospitals, and they all had severe optic nerve damages on admission. Second, although it did not affect our diagnosis of IH, the ICP in a portion of patients was recorded as over 330 mmH₂O, for it exceeded the maximal column manometer used in the neurology departments. Third, for the convenience of patients, the long-term follow-ups were mainly conducted by doctors from local hospitals. These results of ophthalmological examinations were mainly collected and transmitted via the Internet, and the questionnaires were performed by Internet or telephone, which might cause bias. Fourth, in the present study, we mainly evaluated visual functional parameters such as BCVA and VFs, as well as fundoscopy findings, other structural or functional assessments such as visual-evoked potentials, quantitative OCT, and recanalization of sinus in MRI/MRV were not conducted, which will be worth investigation in future research. Fifth, due to poor visual acuity or lack of compliance, 11 patients failed to undergo preoperative and postoperative VFs examinations. Finally, this was a retrospective, nonrandomized, noncomparative study with a relatively small sample size. In the future, a large-scale prospective, randomized, controlled trial comparing different treatment options for CVST-induced visual loss is required.

Conclusively, progressive visual loss remains a big challenge in CVST, which deserves more attention from neurologists, ophthalmologists, and neurosurgeons. Considering its favorable efficacy and safety, ONSF can be considered an important approach to resolving progressive visual loss in CVST combined with anticoagulation and ICP-lowering therapy.