Exophthalmos following mechanical thrombectomy for anterior circulation stroke: A retrospective study and review of literature

Abstract

Background

Anecdotal cases of exophthalmos after acute mechanical thrombectomy have been described. We sought to estimate the incidence in a large cohort of patients with acute anterior circulation stroke treated with mechanical thrombectomy. Secondarily, we aimed to evaluate the underlying mechanism and to differentiate it on imaging from other pathology with similar clinical orbital features.

Methods

Between November 2016 and November 2018, we performed a retrospective single-center study of 250 patients who underwent anterior circulation mechanical thrombectomy. Development of exophthalmos was independently evaluated by two readers on preprocedure and 24-h postprocedure non-contrast cerebral CT.

Results

In the mechanical thrombectomy cohort, six individuals (2.4%) developed interval ipsilateral exophthalmos at 24 h. Of these, at least two patients developed clinical symptoms. There was almost perfect agreement between assessments of the two readers (Cohen’s kappa = 0.907 (95% confidence interval: 0.726, 1.000)). In two patients, there was delayed ophthalmic artery filling on digital subtraction angiography. None of the patients had features of a direct carotid-cavernous fistula.

Conclusions

Exophthalmos is not uncommon after mechanical thrombectomy (2.4%). The underlying mechanism is difficult to confirm, but it is most likely due to orbital ischemia from hypoperfusion or distal emboli.

Introduction

A meta-analysis of five major randomized controlled stroke trials showed that endovascular mechanical thrombectomy (MT) benefits most individuals with acute ischaemic stroke caused by proximal anterior circulation occlusion.¹ The consequent increase in number of patients treated with MT has led to the recognition of previously unreported procedural complications. A recent publication by Brinjikji et al. addressed an individual case of exophthalmos post-MT.² In this case, orbital findings were attributed to orbital infarction syndrome (OIS), an entity defined as ischemia of intraocular and intraorbital structures including orbital fat, optic nerve and extraocular muscles.³ Post-MT exophthalmos has also been described due to a direct carotid-cavernous fistula (CCF) in a small case series.⁴ Because this pathology is limited to occasional case reports and small case series, there is currently a lack of evidence on the prevalence and etiology. With expanded use of MT, we believe it is possible that this pathology is more frequent than previously expected and often gets misinterpreted. The purpose of this retrospective case series is therefore three-fold: to estimate the incidence of the development of exophthalmos in a large cohort of people with acute anterior circulation stroke treated with MT; to evaluate the underlying mechanism for this phenomenon; and finally to differentiate it on imaging from other pathologies with similar clinical orbital features.

Material and methods

This study was approved by our local institutional review board (CE 18.357). The board waived the need for patient consent.

We performed a retrospective single-center cohort study at a high-volume tertiary stroke center with 250 consecutive individuals treated with endovascular MT for anterior circulation strokes (February 2017–November 2018). The MT cohort consisted of 51.2% male and 48.8% female individuals. Median individuals’ age was 69 years (interquartile range 20, 61–81). Our local practice is to perform thrombectomy under monitored local anesthesia with conscious sedation as needed.

Exophthalmos was independently evaluated on preprocedure and 24-h postprocedure non-contrast cerebral CTs by two neuroradiologists with 5- and 25 years’ experience. Readers were blinded to intervention and imaging timing (pre- versus postprocedure). Pre -and post-treatment scans were manually postprocessed by CT technicians. Axial reconstructions were made at the interzygomatic line in which the orbital lens was best seen.

We used a modified concept of Campi et al.⁵ to assess for exophthalmos. First, an interzygomatic line was drawn on the axial view image in which the lens was best seen (Figure 1(a)). Next, a perpendicular line was drawn from the inner retinal surface to the interzygomatic line. The length of this perpendicular line to the interzygomatic line was taken as the primary measurement. Positive development of exophthalmos was defined as an increase of at least 3 mm measured between posterior globe and interzygomatic line between the preprocedure and postprocedure scan. An MRI scan of the orbits was not part of the standard workup in patients with postprocedural exophthalmos.

Figure 1.

Non-contrast axial computed tomography (CT). Interzygomatic line (*) used to determine the degree of exophthalmos (**).

Patient records were reviewed for ophthalmological symptoms and modified Rankin Score (mRS) at three months. Guiding catheter position, intracranial occlusion site and final revascularization result using ‘Thrombolysis In Cerebral Infarction’ (TICI) score were determined from angiographic images. A dedicated ECA angiogram was not part of the standard protocol at the end of the cerebral angiogram. Contrast enhancement in ophthalmic and external carotid arteries and superior ophthalmic vein (SOV) size were evaluated on baseline CT-angiogram (CTA).

Cohen’s kappa (95% confidence interval) was used to test inter-rater reliability⁶ between the two readers, with the following kappa value interpretation (and level of agreement): 0–0.20 (none), 0.21–0.39 (minimal), 0.40–0.59 (weak), 0.60–0.79 (moderate), 0.80–0.90 (strong), above 0.90 (almost perfect).

Results

In the MT cohort, six individuals (2.4%) developed exophthalmos (Table 1). There was almost perfect agreement between assessments of acute exophthalmos by the two readers (Cohen’s kappa = 0.907 (95% confidence interval: 0.726, 1.000)).

Table 1.

Six out of 250 thrombectomy patients (2.4%) showed imaging features of exophthalmos on 24-h non-contrast CT brain. Two patients had orbital symptoms.

Patient	Side	Occlusion	TICI	Exophthalmos increase (mm)	Orbital symptoms	mRS at discharge	SOV size
1	Right	M1	2B	4	Yes	2	Normal
2	Right	ICA terminus	2A	4	Yes	6	Normal
3	Left	ICA terminus	1	4	No	5	Normal
4	Left	ICA + M1	1	6	No	5	Normal
5	Left	M1	3	3	No	3	Normal
6	Left	ICA + M1	2A	4	No	4	Normal

TICI: Thrombolysis In Cerebral Infarction, mRS: modified Rankin Scale. SOV: superior ophthalmic vein.

Of the six cases, five were male and one was female, median age 72 years (age range 56–82). Two patients had tandem (cervical and intracranial), two carotid terminus and two M1 segment occlusion. Four occlusions were on the left and two on the right. Recanalization success was variable: two TICI 1, two TICI 2A, one TICI 2B and one TICI 3. In all six cases, the guiding catheter was placed distal to the carotid bifurcation in the cervical segment of the internal carotid artery. On prethrombectomy CTA, ophthalmic artery contrast opacification was absent in two cases. None of the patients had an SOV dilatation on the pre- or postprocedure scan.

Two out of six patients had orbital symptoms on the side of the thrombectomy documented in their medical files.

The first patient was treated for a right M1-segment occlusion and had a TICI 2B result after MT. One-hour postprocedure, the patient reported ipsilateral orbital pain, chemosis and diplopia. A CT and MRI scan performed showed right-sided exophthalmos, extra-ocular muscle swelling and intraconal fat stranding (Figure 2). The ophthalmologist measured an elevated right eye intra-ocular pressure of 36 mm Hg (normal 12–22). The patient was empirically treated with 5 mg of prednisone IV daily for one week and self-reported improvement within days. The mRS at discharge was 2 with no residual ophthalmological symptoms.

Figure 2.

A patient treated for a right M1 occlusion developed right-sided progressive exophthalmos post-MT. Axial pre- (a) and 24-h post-MT (b) non-contrast CT brain showing development of right-sided exophthalmos (*). Left side remains unchanged (**). Axial (c) and coronal (d) orbital T2-weighted fat saturation MRI sequence showing right-sided thickening and increased extra-ocular muscles signal (white arrows) and intraconal fat stranding (*)

A second intubated patient was treated for a right carotid terminus occlusion and had a TICI 2A result after MT. The right-sided exophthalmos was noted immediately after the MT. The ophthalmologist measured an elevated right eye intra-ocular pressure of 40 mm Hg and the patient was put on 5 mg of Prednisone IV daily. Unfortunately, the patient died 72 h post-MT.

In the MT cohort, there were eight patients with bilateral mild (3–5 mm) exophthalmos on preprocedure scan, which remained stable on 24-h CT. None of these patients were clinically symptomatic.

Discussion

In this single-center study of 250 patients, we investigated post-MT exophthalmos incidence for acute anterior circulation stroke. In the MT group, six individuals (2.4%) developed ipsilateral exophthalmos on 24-h postprocedure CT. Of these, at least two had clinical orbital symptoms. It is, however, very well possible that symptoms were underreported, as the four ‘asymptomatic’ patients had a global aphasia.

A first cause of post-MT exophthalmos is OIS, a very rare ischemic entity only reported in case reports and small case series. A wide variety of etiologies for OIS-like presentations has been described, including orbital cellulitis and mucormycosis, acute perfusion failure secondary to common carotid artery (CCA) occlusion, vasculitides such as giant cell arteritis as well as iatrogenic trauma from craniotomies.^7–9 OIS can also be a result of hypoperfusion in the setting of CCA dissection, atherosclerosis or ligation, affecting ophthalmic and orbital external carotid artery (ECA) supply.^10,11 Two out of six exophthalmos patients had indeed no ophthalmic artery contrast opacification on baseline CTA with delayed filling on DSA, indicating an ischemic mechanism of slow flow or hypoperfusion. A possible mechanism for this is occlusion of the ophthalmic artery during the MT stent retrieving maneuver.

Another possibility is multiple proximal emboli as a result of the MT stent retrieving. A recent case report by Brinjikji et al. described OIS post-MT secondary to emboli.² Our positive post MT exophthalmos cases are, however, very distinct. Brinjikji et al. described OIS secondary to complete occlusion from all orbital branches (internal maxillary artery, ophthalmic artery and superficial temporal artery). The combination of these angiographic findings was not present in any of the positive cases in our case series. Improper placement of the balloon-guiding catheter beyond the carotid bifurcation may contribute to these emboli.^12,13 However, in all positive exophthalmos patients, the balloon-guiding catheter was placed in the ICA, which makes emboli less likely.

A second cause of post-MT exophthalmos is a direct carotid-cavernous fistula (CCF). In this disease, there is shunting between branches of the internal carotid artery with the cavernous sinus. Patients typically exhibit similar symptoms as in OIS, but imaging will demonstrate an enlarged superior ophthalmic vein and early cavernous sinus opacification on the postprocedural scan. A direct CCF as a complication of MT has been described in rare cases.⁴ In our case series, none of the positive exophthalmos patients had an enlarged SOV. Recognizing the absence of an enlarged SOV in a patient with post-MT exophthalmos avoids an unnecessary diagnostic cerebral angiography, as a direct CCF can be excluded from the differential diagnosis.

A final cause would be increased intraocular pressure due to the large infarct size with parenchymal distortion at the level of the superior orbital fissure and/or optic canal. However, in none of the positive patients this finding was present.

Despite that our study is retrospective, the two readers had almost perfect level of agreement. Also, patient number was large, and imaging assessment was blinded to treatment and timing. There may be an evaluation bias, as readers may have recognized postprocedure scans by signs of ischemia from a stroke.

Another limitation is the lack of co-registration of the pre -and post-treatment CT scans. In our study, postprocessing was manually performed by the CT technicians. Unfortunately, we did not have imaging software at our disposal to do this automatically.

Summary

Exophthalmos is not uncommon after MT (2.4%). The mechanism is most likely orbital ischemia due to hypoperfusion or distal emboli, although rare cases of post-MT direct CCF have been reported.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Ethical approval

Ethics approval for this study was obtained by Centre Hospitalier de l’Université de Montréal (CHUM). Number: CE 18.357.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article. of this article.

ORCID iDs

D Volders https://orcid.org/0000-0003-4074-8702

G Jacquin https://orcid.org/0000-0003-4362-1571