Wong LW, Lam WY, Au SCL. Systematic review and risk factor analysis of post-vitrectomy silicone oil migration to the central nervous system. World J Exp Med 2025; 15(3): 106429 [DOI: 10.5493/wjem.v15.i3.106429]
Corresponding Author of This Article
Sunny Chi Lik Au, Chief Physician, Research Fellow, Department of Ophthalmology, Tung Wah Eastern Hospital, 9/F, MO Office, Lo Ka Chow Memorial Ophthalmic Centre, No. 19 Eastern Hospital Road, Causeway Bay, Hong Kong 999077, China. kilihcua@gmail.com
Research Domain of This Article
Ophthalmology
Article-Type of This Article
Systematic Reviews
Open-Access Policy of This Article
This article is an open-access article which was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/
Author contributions: Wong LW drafted the manuscript; Wong LW and Lam WY analyzed the data; Wong LW and Au SCL acquired the data; Lam WY revised the manuscript; Au SCL designed the research study; all authors had full access to the data, contributed to the study, read and approved the final version for publication.
Conflict-of-interest statement: All authors have disclosed no conflicts of interest. This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
PRISMA 2009 Checklist statement: The authors have read the PRISMA 2009 Checklist, and the manuscript was prepared and revised according to the PRISMA 2009 Checklist.
Open Access: This article is an open-access article that was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution NonCommercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: https://creativecommons.org/Licenses/by-nc/4.0/
Corresponding author: Sunny Chi Lik Au, Chief Physician, Research Fellow, Department of Ophthalmology, Tung Wah Eastern Hospital, 9/F, MO Office, Lo Ka Chow Memorial Ophthalmic Centre, No. 19 Eastern Hospital Road, Causeway Bay, Hong Kong 999077, China. kilihcua@gmail.com
Received: February 26, 2025 Revised: March 20, 2025 Accepted: April 7, 2025 Published online: September 20, 2025 Processing time: 167 Days and 18.2 Hours
Abstract
BACKGROUND
Silicone oil (SiO) migration to the central nervous system (CNS) is a rare complication of SiO tamponade after vitreo-retinal surgeries, it could masquerade hemorrhage on computed tomography neuro-imaging. Only limited cases were reported in the literature, certain intra-operative and post-operative ocular risk factors might contribute to the different extend of SiO migration in the CNS.
AIM
To study the risk factors for cerebral ventricular migration (CVM) on top of visual pathway migration (VPM).
METHODS
Conforming to the preferred reporting items for systematic reviews and meta-analyses guidelines, literature searches on PubMed, MEDLINE, EMBASE were performed on June 1, 2024. Publications on SiO migration to CNS were included in this review. Non-English articles, and studies without neuro-imaging of the CNS were excluded. Patient demographics, SiO filled eyes' ocular characteristics and vitrectomy surgical details were extracted from included studies in this review. VPM and CVM were assigned as group 1 and group 2 respectively. Fisher's exact tests, Mann-Whitney U tests and binary logistic regression were performed.
RESULTS
Total 68 articles were obtained after searches, 48 publications were included for analysis. Total 54 SiO filled eyes were analyzed. Post-vitrectomy intraocular pressure (IOP) was found to be significant in both Mann-Whitney U test (P = 0.047) and binary logistic regression (P = 0.012). Diabetic was found to be significant in binary logistic regression (P = 0.037), but at borderline risk for CVM in Fisher's exact test (P = 0.05). Other significant factors include longer SiO tamponade time (P = 0.002 in Fisher's exact test) and visual acuity (P = 0.011 in binary logistic regression). Optic nerve atrophy or disc cupping (P = 1.00, P = 0.790) and congenital optic disc anomalies (P = 0.424) were all with P > 0.05.
CONCLUSION
SiO migration to CNS is rare with limited case reports only. Our analysis of the existing literature demonstrated higher post-vitrectomy IOP was associated with CVM, followed by patients’ diabetic status, longer SiO tamponade time and visual acuity. Optic nerve atrophy, disc cupping and congenital optic disc anomalies were not associated. Modifiable risk factors of post-vitrectomy IOP and SiO tamponade time should be closely monitored by vitreo-retinal surgeons. Lower IOP target post-vitrectomy and earlier SiO removal surgeries should be arranged.
Core Tip: It is interesting to note that silicone oil (SiO) migration to optic chiasm causes visual field loss or deteriorated visual acuity in the contralateral eye. And timely treatment can reverse such complications. Cerebral ventricular migration (CVM) of SiO are largely asymptomatic, commonly being diagnosed on imaging for other medical issues. Most treatments are mainly conservative, especially as when it is asymptomatic. However, CVM was occasionally misrecognized as intraventricular haemorrhage. Healthcare providers should pay attention to this rare but possible event as this misidentification could potentially delay emergency treatments in stroke patients.
Citation: Wong LW, Lam WY, Au SCL. Systematic review and risk factor analysis of post-vitrectomy silicone oil migration to the central nervous system. World J Exp Med 2025; 15(3): 106429
Silicone oil (SiO) is commonly used in complex vitreoretinal surgeries, including tractional retinal detachment, rhegmatogenous retinal detachment accompanied by proliferative vitreoretinopathy, giant retinal tears, and traumatic retinal detachment[1], instead of sulfur hexafluoride (SF6) or perfluoropropane (C3F8) for its long-duration tamponade effect on the retina as it remains in the vitreous cavity. Additionally, it is used when patients require air travel (not fit for SF6 or C3F8) shortly after the surgeries, or have difficulties in posturing post-operatively for long duration[2]. Despite the many reported advantages of SiO, various accounts of complications, with notable gap in long-term safety data, manufacturing variability and purity of SiO products delayed approval of intravitreal use of SiO by the United States Food and Drug Administration (FDA) until 1994. Common complications include cataract formation, emulsification, increased intraocular pressure (IOP), keratopathy, and visual loss[1,2].
Various forms of SiO migration were also reported, such as subconjunctival and orbital. Although being under-recognized, retrolaminar migration to the optic nerve and chiasm may be commoner than thought and potentially symptomatic. It is generally assumed that SiO infiltrates the optic nerve directly through the lamina cribrosa (along the axis of the entering axons) facilitated by post-operative IOP elevations. The overall rate of optic nerve infiltration in previous studies of enucleated eyes was 20%[3]. However, not every patient with SiO in eye would be scanned regularly with computed tomography (CT) covering the optic nerve region, and most patients do not require enucleation, so the true prevalence is likely to be under-estimated. SiO might even migrate further to the subarachnoid space if there is a defect, such as rupture in the periphery of the lamina cribrosa[4], congenital optic disc pits[5] or natural gaps where the central retinal artery and vein pass through the optic nerve[3]. Very rarely, the subarachnoid space of the optic nerve communicates with the subarachnoid space of the brain and results in intraventricular SiO deposition[6].
SiO migration to the central nervous system (CNS) is a rare complication of SiO tamponade after vitreo-retinal surgeries, limited cases were reported. Our systematic review followed the preferred reporting items for systematic reviews and meta-analyses (PRISMA) guidelines. We aim to conduct analysis of published literature in order to better understand the reported risk factors and symptoms for visual pathway migration (VPM) vs cerebral ventricular migration (CVM).
MATERIALS AND METHODS
Literature search
Conforming to the PRISMA guidelines, systematic literature search (search date June 1, 2024) of the databases EMBASE, MEDLINE and PubMed were performed using the search terms “silicone oil migrat” and “central nervous system” or “brain” or “optic nerve” or “optic chiasm” or “subarachnoid space” or “cerebral ventricles” or “third ventricle” or “lateral ventricle” or “fourth ventricle” or “sellar”. Details of the search strategy are listed out in Supplementary Tables 1 and 2. Publications on SiO migration to CNS from 1994 (when SiO was approved by United States FDA) to 2024 were included in this review. Non-English articles, and studies without neuro-imaging of the cerebral ventricles were excluded. Inclusion and exclusion criteria were listed out in Table 1. The PRISMA flow diagram is presented in Figure 1.
Figure 1 The preferred reporting items for systematic reviews and meta-analyses flow diagram.
Table 1 Inclusion and exclusion criteria of this systematic review.
Criteria
Inclusion criteria
Publications on silicone oil migration to central nervous system
Publications from 1994 to June 1, 2024
All types of publications, including case reports, case series, retrospective case control, cohort studies, randomized controlled trials
Exclusion criteria
Non-English language publications
Studies without neuro-imaging of the cerebral ventricles
Statistical analysis
Patient demographics (age, gender, diabetic status, visual acuity), presenting symptoms (headache, nausea, dizziness, loss of consciousness, mental change, painful eye, visual field loss and visual change), SiO-filled eyes' ocular characteristics (presence of congenital disc pit, optic disc cupping, buckle, glaucoma drainage device) and vitrectomy surgical details (SiO emulsification, SiO endotamponade time, post-operative IOP) were extracted. Visual acuities were converted to LogMAR for analysis. Since most patients had low vision, finger counting, hand movement, light perception, and no light perception were assigned values of 1.7, 2.0, 2.3, and 3.0 LogMAR, respectively[7]. VPM and CVM were assigned as group 1 and group 2 respectively. VPM (group 1) and CVM (group 2) were analysed using Fisher's exact test for categorical variables, Mann-Whitney U test for non-parametric unpaired data, and binary logistic regression for association quantification with Statistical Package for the Social Sciences v29. A two-tailed P value of less than 0.05 was considered statistically significant.
RESULTS
The database search yielded 126 records. After removing duplicates, 68 articles were obtained. Excluding 2 non-English articles, 66 papers were assessed for full text review. These studies were published between the years 1994 and 2024. All were case reports or series. No cohort study, retrospective case control, nor randomized controlled trial were found. Among these, 16 were excluded due to inadequate individual case data, and 2 were excluded for lacking cerebral ventricle imaging. Total 48 studies were included into the systematic review, with 54 patients and 54 eyes being reviewed[3-6,8-51].
Demographics in both groups
The median age of patients was 58 years (range, 27-93 years) and 31 (57.4%) were men. Most common comorbidity reported was diabetes mellitus (25 patients, 46.2%). Only 34 patients had their visual acuity reported after operation. Total 27 patients (79.4%) had no light perception, 3 patients (8.8%) had light perception, 1 patient (2.9%) could count fingers. The remaining 3 patients had poor visual acuity of logMAR value 0.70, 0.92 and 1.3 respectively.
Patients presented with different ocular characteristics, including optic nerve atrophy or cupping (n = 11), congenital optic pit (n = 1), buckling (n = 3), and glaucoma drainage device implanted (n = 3).
Indication for intraocular SiO placement included tractional retinal detachment due to proliferative diabetic retinopathy (n = 24), rhegmatogenous retinal detachment (n = 21), cytomegalovirus retinitis (n = 2), trauma (n = 1), terson syndrome (n = 1), and not stated (n = 5).
Post-operatively, the average SiO endotamponade time was 45.8 months (range, 1 day to 20 years). The mean post-operative IOP was 42.4 mmHg (range, 14-80 mmHg), with 17 of them requiring treatment. Total 3 eyes (5.6%) had SiO emulsification reported.
Presenting symptoms in both groups
Most common presenting symptoms after SiO placement was headache (n = 20), followed by loss of consciousness (n = 10), mental change (n = 6), painful eye (n = 4), dizziness (n = 9), nausea (n = 4), visual field loss (n = 1).
SiO migration to visual pathway
VPM was allocated in group 1 for analysis. It was reported in 11 eyes with at least optic nerve migration of 11 patients. One got migration up to the supresellar region, another one got subarachnoid space along the optic nerve migration. Five eyes were with optic chiasm migration.
Median age was 51 years, 6 men and 5 women. The average post-operative IOP was 31.9 mmHg and mean SiO endotamponade time was 10.3 months. Presenting symptoms were diverse, most present with headache (n = 3), others include dizziness (n = 2), painful eye (n = 2), visual change (n = 2), loss of consciousness (n = 1), mental change (n = 1) and visual field loss (n = 1). Nine patients were managed conservatively without further deterioration. Specific treatment to remove SiO in the visual pathway was reported in two case reports. One patient underwent optic nerve sheath decompression and SiO drainage from the subarachnoid space within the optic nerve sheath[8], but vision showed no improvement. Another patient[6] (the left eye studied) presented with contralateral (right eye) visual field loss, with the left optic nerve slit open at the transition of the optic chiasm. The intra chiasmatic and intraneural SiO was removed by smooth suction. One month later, a regression of the visual field defect in the right eye was observed. The IOP in the left eye remained normal without treatment after surgery.
SiO migration to cerebral ventricles
CVM was allocated in group 2 for analysis. It was reported in 43 eyes of 43 patients. Diagnosis of SiO migration was all by neuro-imaging including CT and magnetic resonance imaging (MRI) scans. The migration inevitably started from the globe, and many followed the optic nerve, and SiO is often present at multiple locations with movement from different postures. Concerning the final destination of migration, the distribution of migration were lateral ventricles only (n = 33), lateral and third ventricles (n = 2), lateral and fourth ventricle (n = 3), lateral, third and fourth ventricle (n = 4) and optic tract (n = 1).
Median age was 59.5 years, 25 men and 18 women. The average post-operative IOP was 49.1 mmHg and mean SiO endotamponade time was 67.3 months. Again, the majority presented with headache (n = 17), followed by loss of consciousness (n = 9), visual change (n = 8), dizziness (n = 7), mental change (n = 5), nausea (n = 4), painful eye (n = 2).
Most patients were managed conservatively with symptomatic relief, such as antihypertensive medication for patients experiencing headaches, and control of IOP. The majority of the treatment was not targeted towards intraventricular SiO migration as most recovered without removing the SiO from ventricles. One reported patient had SiO removal from the chiasm rather than the ventricle shows significant improvement in visual acuity[36]. This patient who had SiO in the right eye experienced vision changes. The right eye had no light perception after the SiO procedure, while vision in the left eye declined from 20/20 initially to 20/70 later on. On imaging, the SiO was detected along the optic nerve, optic chiasm, and lateral ventricles. Subsequently, the patient underwent a procedure to remove SiO from the optic chiasm only. Vision in the left eye improved to 20/20 after the surgery. The patient was then given prednisone daily. Six months after stopping the steroid treatment, there was no recurrence of vision loss in the left eye. A follow-up MRI showed a significant reduction of SiO within the chiasm, illustrating the visual deterioration was caused by the SiO migration to the chiasm rather than the ventricles.
Differences in migration
The demographic, ocular characteristic, vitrectomy surgical details and presenting symptoms between group 1 (VPM) and group 2 (CVM) are compared in Table 2. Post-vitrectomy IOP was found to be significant in both Mann-Whitney U test (P = 0.047) and binary logistic regression (P = 0.012). Diabetic was found to be significant in binary logistic regression (P = 0.037), but at borderline risk for CVM in Fisher's exact test (P = 0.05). Other significant factors include longer SiO tamponade time (P = 0.002 in Mann-Whitney U test) and LogMAR visual acuity (P = 0.011 in binary logistic regression). Bonferroni correction was applied to adjust for multiple comparisons in this small sample sized study[52], and these four factors: (1) Post-vitrectomy IOP (P = 0.033); (2) Diabetic status (P = 0.043); (3) SiO tamponade time (P = 0.013); and (4) LogMAR visual acuity (P = 0.004) were statistically significant. Other factors remain P > 0.05, i.e. statistically insignificance. Multivariable model was applied, post-vitrectomy IOP, SiO tamponade time, LogMAR visual acuity, and indication for intraocular SiO placement were found to be confounding variables. After adjustment to these confounding variables, all factors (P > 0.05) tested were not associated with the extend of SiO migration.
Table 2 Details comparison of eyes included in group 1 and group 2, n (%).
There is no direct anatomical communication between the vitreous cavity and cerebral ventricles. Therefore, any additional defect between them is likely to be regarded as a risk factor for SiO migration, such as congenital optic disc pits or natural gaps where the central retinal artery and vein pass through the optic nerve. Indirect route of migration through optic nerve sheath fenestration might be possible. However, optic nerve atrophy, cupping, or congenital optic disc pit were found to be insignificant in our systematic review. This might be a result of small sample size from limited mention of such among the included publications, thus a potential type II error.
It is also proposed that SiO infiltration is facilitated by post-operative IOP elevations and a long endotamponade time. Under these two driving forces, SiO is thought to migrate directly through the lamina cribrosa (a mesh-like structure at the optic nerve head that forms a pressure barrier and allows nerve fibers to leave the eye and form the optic nerve) along the axis of the entering axons, or through any rupture in the periphery of the lamina cribrosa[44,53,54]. In this study, post-vitrectomy IOP was found to be significant in both statistical tests (Mann-Whitney U test and binary logistic regression) and longer SiO tamponade time. Post-operative IOP may be related to SiO tamponade duration, thus a potential confounding factor, affecting the reliability of regression results. Isolated elevation of IOP may have a role to play in SiO migration. Notably, the mean IOP in both groups were above the physiological 21 mmHg, yet the data from individual studies on the diagnosis of clinical glaucoma were not complete, hence glaucomatous optic neuropathy could not be concluded. Modifiable risk factors of post-vitrectomy IOP and SiO tamponade time should be closely monitored by vitreo-retinal surgeons. Lower IOP target post-vitrectomy and earlier SiO removal surgeries should be arranged. Very rarely, the subarachnoid space of the optic nerve communicates with the subarachnoid space of the brain, resulting in intraventricular SiO deposition.
Diabetic was found to be significant in binary logistic regression, but at borderline risk for CVM in Fisher's exact test. Hypothetic association might be the surgical technique employed in tractional retinal detachment surgery compared to rhegmatogenous retinal detachment surgery. Diabetic retinopathy subjects who required vitrectomy surgeries were likely suffering from macular-off tractional retinal detachment. Tractional membranes must be relieved intra-operatively to flatten the retina and macula before SiO injection to the vitreous cavity. The peeling process might elicit some breaks along the retina, especially over the posterior pole and adjacent to the optic discs where fibrous proliferation usually occurred, causing defects for subsequent SiO migration. This might also explain why poor post-operative visual acuity is one of the risk factors that was found to be significant, as tractional retinal detachment with underlying uncontrolled diabetes usually yields worse results than rhegmatogenous retinal detachment post-operatively.
Presenting symptoms and management
VPM and CVM are usually incidental findings in neuro-imaging during evaluation of non-specific symptoms such as headache and loss of consciousness. Most of these presenting symptoms have an alternative medical diagnosis, such as hypertension and delirium. Therefore, treatments are mostly symptomatic and supportive. However, the site of migration may predict the success of symptom relief by SiO removal. Two case reports suggested removing SiO from the optic chiasm could improve vision[5,36], while similar efficacy was not observed in SiO from optic nerve[8]. This is likely due to a relief on the direct pressure compressing on the chiasm per se. It is not needed to remove SiO from the ventricles as it mainly remains asymptomatic.
Although it stays in the ventricle without causing much symptoms, the real challenge is usually in differentiating SiO from other conditions like hemorrhage[27,32,49], colloid cysts, aneurysms, or subarachnoid blood, depending on their location. To differentiate SiO from hemorrhage on CT, it is important to be aware that the lower physical density of oil means that it floats in water, in contrast to blood that will sink to gravity dependent area[55]. Therefore, it is important to combine the imaging findings with patients’ previous ophthalmic surgical history. Moreover, SiO usually presents as radiologically globular and non-dependent hyperdense lesions on non-contrast CT. Interval scan or postural change during imaging could show a migration of SiO within ventricles, which differentiate it from hemorrhage or aneurysms. Most importantly, a good post-operative IOP control is always needed after ocular surgery not only to prevent glaucomatous change or causing headache in patients, and to minimize the risk of potential SiO migration as suggested by this study.
Limitations
Despite our systematic review was conducted according to the PRISMA guidelines, it was not registered under the International Prospective Register of Systematic Reviews PROSPERO. The major limitation of this systematic review is the limited number of studies to be reviewed as VPM and CVM are rare complications, lacking high-quality cohort or controlled studies. And since most case reports are incidental findings, tend to focus on rare or severe cases, the true prevalence is likely to be underestimated. This could overestimate the association of certain factors (such as high IOP) while underestimating the role of other potential factors. Due to the rarity of SiO migration to the CNS, the existing data are limited (only 54 patients), leading to insufficient statistical power that may not only fail to accurately reflect the true risk factors, but also induce the risk of exaggerating statistical significance and false positives.
Publication bias is a significant limitation of this systematic review, as case reports are more likely to be published if they report unusual or extraordinary cases. Our systematic review is of exploratory nature, and attempted to explore the association of various risk factors for the different extend of SiO migration to CNS. Moreover, the risk factors examined were not consistently documented in all the chosen studies, so they are assumed to be absent. This could introduce the potential for misleading comparisons, as these factors might indeed be present in patients but were not consistently reported in the literature. In addition, quite a large proportion of patients were of non-numerical visual acuity, and numbers were assigned for finger counting, hand movement, light perception and no light perception. Measurement bias could be introduced. External validity of our study is limited, future multi-center registry studies to collect larger samples are definitely needed.
CONCLUSION
SiO migration to CNS is rare with limited case reports only. Our analysis of the existing literature demonstrated higher post-vitrectomy IOP was associated with CVM, followed by patients’ diabetic status, longer SiO tamponade time and visual acuity. However, their significances were not observed after adjustment to confounding variables in our multivariable model analysis. Modifiable risk factors of post-vitrectomy IOP and SiO tamponade time should be closely monitored by vitreo-retinal surgeons. Lower IOP target post-vitrectomy and earlier SiO removal surgeries should be arranged. CVM was occasionally misrecognized as intraventricular hemorrhage. Healthcare providers should pay attention to this rare but possible event as this misidentification could potentially delay emergency treatments in stroke patients. Thorough medical history taking, comprehensive clinical evaluation, combining with radiological imaging are essential to arrive at the correct diagnosis. Given the current literature is lack of cohort or controlled studies, and with just small number of cases published, future researches of large-scale case control and prospective cohort studies with multi-center registries on SiO migration are needed to validate the risk factors.
Footnotes
Provenance and peer review: Invited article; Externally peer reviewed.
Peer-review model: Single blind
Specialty type: Medicine, research and experimental
Country of origin: China
Peer-review report’s classification
Scientific Quality: Grade B, Grade B, Grade C, Grade D
Novelty: Grade A, Grade B, Grade C, Grade E
Creativity or Innovation: Grade B, Grade B, Grade C, Grade E
Scientific Significance: Grade B, Grade B, Grade B, Grade C
P-Reviewer: Glumac S; Salimi M; Wu YB S-Editor: Luo ML L-Editor: A P-Editor: Zheng XM
Chen JX, Nidecker AE, Aygun N, Gujar SK, Gandhi D. Intravitreal silicone oil migration into the subarachnoid space and ventricles: A case report and review of literature.Eur J Radiol Extra. 2011;78:e81-e83.
[PubMed] [DOI] [Full Text]
Madireddy NK, Polu AR, Vishwanath S. Misdiagnosis of intraventricular hemorrhage in a patient with intraocular silicone: a radiographic caveat (3035).Neurology. 2021;96.
[PubMed] [DOI] [Full Text]
