© Springer Nature Switzerland AG 2019
Andrew G. Lee, Alexandra J. Sinclair, Ama Sadaka, Shauna Berry and Susan P. Mollan (eds.)Neuro-Ophthalmologyhttps://doi.org/10.1007/978-3-319-98455-1_11

11. Imaging of Oculomotor (Third) Cranial Nerve Palsy

Michael S. Vaphiades1  , Martin W. ten Hove2  , Tim Matthews3  , Glenn H. Roberson4   and Alexandra Sinclair5, 6  
(1)
Departments of Ophthalmology, Neurology and Neurosurgery, University of Alabama at Birmingham, Birmingham, AL, USA
(2)
Department of Ophthalmology, Queen’s University, Kingston, ON, Canada
(3)
Department of Ophthalmology, University Hospital Birmingham, Birmingham, UK
(4)
Department of Radiology, University of Alabama at Birmingham, Birmingham, AL, USA
(5)
Metabolic Neurology, Institute of Metabolism and Systems Research, College of Medical and Dental Sciences, University of Birmingham, Birmingham, UK
(6)
Department of Neurology, University Hospitals Birmingham NHS Foundation Trust, Queen Elizabeth Hospital, Birmingham, UK
 
 
Michael S. Vaphiades (Corresponding author)
Email: mvaphiades@uabmc.edu
 
Martin W. ten Hove
Email: tenhove@queensu.ca
 
Tim Matthews
Email: tim.matthews@uhb.nhs.uk
 
Glenn H. Roberson
Email: radghr@uab.edu
 
Alexandra Sinclair
Email: a.b.sinclair@bham.ac.uk

Keywords

Magnetic resonance angiography (MRA)Computed tomography angiography (CTA)Intra-arterial digital subtraction angiography (DSA)

Key Points

  1. 1.

    Although in the past it was recommended that vasculopathic patients with isolated, pupil spared, third nerve palsy may be observed, in modern clinical practice, most patients presenting with an oculomotor (third) cranial nerve palsy should probably receive some form of neuroimaging study.

     
  2. 2.

    The main driver for the recommendation for imaging is that a significant number of patients with a ruptured intracranial aneurysm resulting in subarachnoid hemorrhage had an oculomotor palsy prior to rupture and earlier diagnosis and treatment of aneurysm can make a difference in morbidity and mortality.

     
  3. 3.

    Although catheter digital subtraction angiography (DSA) remains the definitive “gold standard” for imaging cerebral aneurysms (including those producing an oculomotor palsy), less invasive angiographic techniques may be available in the initial evaluation armamentarium.

     
  4. 4.

    In most centers, computed tomography angiography (CTA) has supplanted DSA in many cases as the initial imaging study for intracranial aneurysms. Also, conventional computed tomography (CT) (without contrast) is useful for evaluating for subarachnoid hemorrhage in the acute setting.

     
  5. 5.

    Magnetic resonance angiography (MRA) remains a powerful alternative or complementary study to both DSA and CTA when imaging intracranial aneurysms especially when there are artifacts or contraindications associated with these studies. In addition, conventional magnetic resonance imaging (MRI) (with and without contrast) is superior to both CT/CTA and DSA for the evaluation of non-aneurysmal causes of third nerve palsy.

     

Introduction

The initial evaluation and management of a patient with an isolated oculomotor cranial nerve palsy is challenging and a high stakes encounter due to the possibility of aneurysmal compression as the etiology. There is still debate on which initial mode of neuroimaging should be ordered in these cases including MRI or MRA; CT or CTA (Figs. 11.1 and 11.2); or standard catheter DSA [1]. Clinically, compressive lesions of the oculomotor nerve due to aneurysm, often produce ptosis, a fixed, dilated ipsilateral pupil and a specific pattern of oculomotor nerve related ophthalmoplegia (i.e., a pupil involved third nerve palsy with the eye being “down and out” and with a partial or complete ptosis). There should be intact abduction (normal sixth nerve function) and evidence of incyclotorsion in downgaze (normal fourth nerve function). In contrast, oculomotor cranial nerve palsy from an ischemic mononeuropathy (e.g., diabetes, hypertension) usually demonstrates little or no anisocoria and the pupil is usually reactive in the setting of complete ptosis and ophthalmoplegia (i.e., a pupil spared third nerve palsy). There are important exceptions to these general guidelines however:
  1. 1.

    A third nerve palsy due to an aneurysm at or near the junction of the internal carotid and posterior communicating arteries may initially demonstrate normal pupillary size and reactivity (i.e., “pupil spared” third nerve palsy) in up to 14% of patients (especially with partial somatic involvement) and pupillary involvement may eventually develop in the ensuing 7–10 days in such cases [2, 3]. Thus, pupil sparing cannot be used to exclude the diagnosis of aneurysm, especially in partial third nerve palsies.

     
  2. 2.

    In order to be judged as truly “pupil sparing” the isocoric and reactive pupil of the third cranial nerve must be seen in a setting of complete ptosis and complete involvement of the muscles innervated by the oculomotor nerve (i.e., a complete external ophthalmoplegia for extraocular muscles innervated by the third cranial nerve) [3, 4]. If there is not complete ptosis or if the pattern of oculomotor nerve related ophthalmoplegia is incomplete (i.e., partial palsy) then “all bets are off”.

     
  3. 3.

    In addition, ischemic third nerve palsies have been reported to be “pupillary involving” in only up to 32% of cases [4], and the associated anisocoria may be as great as 2.5 mm [5].

     
  4. 4.

    Pain may be present with both compressive and ischemic causes and therefore may not be very helpful in distinguishing between these two entities [6].

     
../images/436228_1_En_11_Chapter/436228_1_En_11_Fig1_HTML.png
Fig. 11.1

Computed tomography angiography: three-dimensional color-rendered angiogram. Note the projection is from above, and the right posterior communicating aneurysm is on the right side of the scan

../images/436228_1_En_11_Chapter/436228_1_En_11_Fig2_HTML.png
Fig. 11.2

Computed tomography angiography. Axial (left) maximal image projection and oblique sagittal (right) views of the right posterior communicating aneurysm. Note relation to cavernous sinus and sella

In clinical practice, most patients presenting with oculomotor cranial nerve palsy, either pupil involving or sparing, receive some form of neuroimaging study because the risk of missing a symptomatic posterior communicating artery (PCom A) aneurysm is very high (Fig. 11.3).
../images/436228_1_En_11_Chapter/436228_1_En_11_Fig3_HTML.png
Fig. 11.3

Axial unenhanced computed tomography scan showing slight hemorrhage in the right paracavernous region

Intracranial Aneurysms

Up to 90% of intracranial aneurysms occur in the circle of Willis [7]. The typical clinical presentation of a ruptured aneurysm with a subarachnoid hemorrhage (SAH) is severe headache (i.e., “worst headache of my life”), loss of consciousness, stiff neck, nausea and SAH may ultimately result in coma or death [8]. Rupture of an intracranial aneurysm with SAH has a mortality of up to 50% [9]. The overall frequency of intracranial aneurysms in the general population is approximately 5%, with a rupture rate for PCom A aneurysm being higher than the anterior circulation aneurysms, at all sizes [10]. Several studies have shown a risk of rupture of 0.52% for aneurysms measuring 7–12 mm in diameter and involving the anterior circulation and 2.9% for aneurysms located in posterior circulation in same size category [11]. One study suggested that the critical size for aneurysm rupture ranges between 4 and10 mm [9]. Another study calculated the rupture rate for unruptured intracranial aneurysms <10 mm in diameter as 0.05% per year [10]. In patients with unruptured intracranial aneurysms of less than 7 mm in diameter who have not had a previous subarachnoid hemorrhage, the rupture rate is about 0.1% per year [11] and for aneurysms >10 mm it approaches 1% per year [10]. If the aneurysm is greater than 25 mm in diameter (e.g., giant aneurysm), the rupture rate is 6% in the first year [10]. Many factors including site, size, and group specific risks are involved in management of these patients with unruptured intracranial aneurysms [11]. The smallest PCom A aneurysm reported to presumably cause a third nerve palsy is 3 mm [12, 13], yet most reports have maintained that a PCom A aneurysm needs to be at least 4 mm to cause a compressive third nerve palsy [12]. Thus, non-invasive initial neuroimaging studies (e.g., contrast enhanced MRA and/or CTA) are vital to detection and management of unruptured intracranial aneurysms and are important alternatives to initial DSA in their detection [14] (Figs. 11.4, 11.5 and 11.6).
../images/436228_1_En_11_Chapter/436228_1_En_11_Fig4_HTML.png
Fig. 11.4

Axial (left) and coronal (right) unenhanced computed tomography scans showing calcification in the wall of a giant right posterior communicating artery aneurysm

../images/436228_1_En_11_Chapter/436228_1_En_11_Fig5_HTML.png
Fig. 11.5

Computed tomography angiography (top left), catheter digital subtraction angiography (top right), and three-dimensional surface-rendered (bottom) images of the giant posterior communicating artery aneurysm

../images/436228_1_En_11_Chapter/436228_1_En_11_Fig6_HTML.png
Fig. 11.6

Three-dimensional (3D) MR angiography. Axial source image (maximal image projection) (left) and sagittal 3D time of flight image (right) showing a partially thrombosed posterior communicating artery aneurysm (arrow)

Digital Subtraction Angiography (DSA)

DSA also known as conventional catheter angiography, utilizes fluoroscopy and iodine-based intravascular contrast material. DSA “subtracts out” images of structures other than blood vessels (e.g., bone) and makes the blood vessels more visible (i.e., digital subtraction). Although modern DSA techniques are associated with a less than 1–2% risk of complications, including stroke [15, 16], this low but not zero risk must be considered in the context of the potential benefits of less invasive initial screening imaging (e.g., CTA or MRA). Although DSA remains the definitive or “gold standard” test for imaging the intracranial and extracranial blood vessels and for detecting cerebral aneurysms, many centers are moving to initial screening with CTA or MRA [15, 17]. In 2010, Thiex et al., evaluated 1715 consecutive patients undergoing DSA and retrospectively assessed them for stroke or transient ischemic attack (secondary to the procedure). In 40 of the patients, a diffusion weighted imaging (DWI) sequence on MRI had been performed within the first 30 days after cerebral angiography. Two patients had punctate areas of restricted diffusion on DWI. Although no stroke or permanent neurologic deficit was seen in any of the 1715 patients, one patient experienced a TIA. Non-neurologic complications without long-term sequelae occurred in nine patients. The risk for neurologic complications related to DSA in this study was less than 1% [18]. DSA is superior to CTA and MRA in that it can detect aneurysms smaller than 3 mm in diameter [19]. The smallest PCom A aneurysm reported to cause a third nerve palsy however is 3 mm [12, 13], with most being at least 4 mm in size [12].

Computed Tomography Angiography (CTA)

Conventional CT and MRI scans have low spatial resolution but much higher contrast resolution whereas DSA has better spatial resolution [20]. CTA uses intravenous (rather than intra-arterial) iodinated contrast and is associated with less morbidity and mortality than DSA. The IV bolus of contrast material is followed by high-speed spiral CT scanning and the patient can be moved through the CT scanner during one breath hold [17]. Shaded surface display (SSD) and maximum intensity projection (MIP) techniques allow improved visualization of potential aneurysms [17, 20]. SSD can show surface anatomy in a 3D type view. Color imaging also helps create a “true life image” on CTA. MIP shows only the blood vessels, without color, and looks more like an MRA source image than an SSD image. Both types of images can be rotated in space. Newer and evolving multidetector technology has improved the scan time to produce these complex images. CTA has good correlation with DSA and exquisite delineation of intracranial anatomy. In one study, CTA detected intracranial aneurysms as small as 3 mm in size [21]. A recent meta-analysis of 45 studies compared CTA with DSA and/or intraoperative findings in patients suspected of having cerebral aneurysms. They found that the diagnostic accuracy of CTA with a 16- or 64-row multi-detector scan was significantly higher than that of single-detector CT (especially in detecting small aneurysms of ≤4 mm in diameter). These authors concluded that CTA may 1 day selectively replace DSA in patients suspected of having a cerebral aneurysm when using a multidetector CT unit [22]. The sensitivity of CTA for the detection of cerebral aneurysms <5 mm can be the same or even higher than that of DSA with equal specificity and high inter-operator reliability in some series [19]. Since CTA is now widely used as a routine primary tool in diagnosis of intracranial aneurysms, there is concern that the radiation dose has been increasing with potential harm to the brain [23, 24]. With this in mind, a prospective study of 294 consecutive patients with spontaneous subarachnoid hemorrhage were randomly assigned to conventional voltage CTA (C-CTA) and low voltage CTA (L-CTA) and analyzed for quality, radiation dose and accuracy of the scan. The authors found no significant differences in sensitivity, specificity, and accuracy between the C-CTA and L-CTA groups and concluded that L-CTA should be recommended as a routine scanning method for the detection of aneurysms because of the lower radiation dose [23]. As revolutionary as CTA has become, there are disadvantages other than radiation dose, for example an aneurysm may be obscured by boney artifact [14]. The advantages of CTA over MRA, include shorter scanning times and better resolution of images. CTA can also be used safely in patients in whom implanted metal objects (such as ferromagnetic shrapnel, older non-MR compatible pacemakers, and older aneurysm clips) which preclude the use of MRI, and is superior to MRI when claustrophobia makes MRI scanning impractical.

Magnetic Resonance Angiography (MRA)

MRA does not utilize ionizing radiation or iodinated contrast material, unlike DSA and CTA. In terms of radiation exposure from CT examinations, one study estimated that of approximately 600,000 abdominal and head CT examinations performed annually in the United States in children under the age of 15 years. They estimated that that 500 of these individuals might ultimately die from cancer attributable to the CT radiation [25]. Granted, the incidence of subarachnoid hemorrhage increases with age (mean age of approximately 50 years), this pediatric radiation data is also important for young adults, especially given that the US population is living longer. Another advantage of MRA is the flexibility of the display (images the blood vessels can be rotated and viewed from different angles). There are three types of MRA techniques: (1) Three (3)-dimensional time of flight (3D TOF) (Figs. 11.7 and 11.8); (2) Two (2)-dimensional time of flight (2D TOF); and (3) contrast-enhanced MRA (CEMRA) [26]. The first two techniques do not require gadolinium contrast, but are dependent upon directional blood flow to generate images [27]. Frequently, all three techniques are used to generate MRA images [27, 28]. CEMRA is used to image both the head and neck vasculature, particularly the aortic arch and great neck vessels. 2D TOF is used to image the carotid arteries and 3D TOF is used to image the intracranial vessels. As indicated, to generate a TOF image, data must be obtained from flow characteristics of blood and a saturation pulse is used to subtract out the veins. Because of blood flow physiology, blood vessel diameter may appear smaller than it actually is, thereby leading to the potential for MRA to “overcall” the degree of vascular stenosis (as in the assessment of the carotid arteries). In this setting, carotid Doppler ultrasonography, and DSA may provide more accurate measurements of the degree of carotid stenosis. In addition, one very important disadvantage of MRA is that despite advances over the past decade, this technique may underestimate the size of intracranial aneurysms, or may fail to detect aneurysms all together, even in patients with characteristic signs such as oculomotor cranial nerve palsy [15, 29]. In fact, patients with oculomotor cranial nerve palsy from PComA aneurysms measuring up to 7 mm have been missed by MRA [1]. One study concluded that only a 2 mm aneurysm seen on catheter-based angiography was missed by MR angiography, yet this is facility dependent and the study was performed at a facility with excellent MRA [30]. The authors also indicated that in their study, symptomatic aneurysms were equal or greater than 4 mm in size [30]. In one study examined 47 patients who had saccular or giant intracranial aneurysms imaged by MRA compared with 19 patients imaged with DSA. The remaining 28 patients, in whom no aneurysm was found, served as a control group. The study concluded that MRA can define the circle of Willis sufficiently to allow detection of intracranial aneurysms as small as 3–4 mm and holds promise as a truly noninvasive screening examination of intracranial vasculature in patients at risk for aneurysms [13]. Another study agreed that CTA and MRA have yielded a detection rate of aneurysms measuring at least 4 mm in diameter approaching the detection rate of DSA [20]. The authors cautioned that to avoid diagnostic mishaps, these imaging studies should be reviewed by at least one neuroradiologist before aneurysm is rejected as the cause or before the patient undergoes DSA [20]. So these imaging studies have, at their best, the capability to detect aneurysms as small as 4 mm but what is the average size of an aneurysm causing a compressive isolated third nerve palsy? In a retrospective review of 417 cases with PCom A aneurysmal third nerve palsies, patients were classified based on the results of the noninvasive neuroimaging obtained at initial presentation. The average size of PCom A aneurysms causing an isolated third nerve palsy was 7.3 mm. This study, like the previous one outlined, also concluded that aside from an accurate history, the training and experience of the interpreting radiologist is probably the most important factor in determining the reliability of a noninvasive scan in patients with isolated third nerve palsies [12]. Another group of investigators evaluated the sensitivity and specificity of MRA in the diagnosis of ruptured and unruptured intracranial aneurysms using a meta- analysis. They found 12 studies from 4 electronic databases on relevant articles published from January 1998 to October 2013. Among the 960 patients assessed, TOF MRA technique was used in most of the studies and showed comparable diagnostic performance to CEMRA. Among studies using 3-Tesla field strength, both sensitivity and specificity were higher compared with studies using lower field strength. They also concluded that MRA is comparable with CTA in diagnosing intracranial aneurysms yet specificity of MRA seems to be slightly lower than the specificity of CTA [31] (Figs. 11.9, 11.10, 11.11 and 11.12).
../images/436228_1_En_11_Chapter/436228_1_En_11_Fig7_HTML.png
Fig. 11.7

Three-dimensional time of flight MR angiography. Image signal intensity is proportional to flow. Vascular compartments have been manually separated

../images/436228_1_En_11_Chapter/436228_1_En_11_Fig8_HTML.png
Fig. 11.8

Three-dimensional time of flight image. Manual separation of components

../images/436228_1_En_11_Chapter/436228_1_En_11_Fig9_HTML.png
Fig. 11.9

Two-dimensional time of flight MR angiography. Signal is proportional to flow and is directionally encoded; only antegrade flow is displayed

../images/436228_1_En_11_Chapter/436228_1_En_11_Fig10_HTML.png
Fig. 11.10

Contrast-enhanced MR angiography with intravenous injection of Gadolinium. Vascular compartments manually separated. Signal intensity is very strong with excellent detail

../images/436228_1_En_11_Chapter/436228_1_En_11_Fig11_HTML.png
Fig. 11.11

Contrast-enhanced MR angiography of the head only

../images/436228_1_En_11_Chapter/436228_1_En_11_Fig12_HTML.png
Fig. 11.12

Sagittal T1 unenhanced MRI scan showing a partially thrombosed aneurysm

Conclusion

In a patient with an acute, isolated, third nerve palsy, non-contrast CT of the head (to exclude subarachnoid hemorrhage) followed by a contrast CTA is generally more sensitive than MRA in the detection of aneurysm at most centers. At their best, both modalities have the capability to detect aneurysms as small as 4 mm in diameter. DSA can detect aneurysms smaller than 3 mm in diameter, yet most reports have maintained that a PCom A aneurysm needs to be at least 4 mm to cause a third nerve palsy. For best results, these imaging studies should be reviewed by at least one neuroradiologist. If an aneurysm is not the only pathology in the differential diagnosis of the third cranial nerve palsy, then the combination of MRA and CTA is more powerful than either technique alone. In addition, conventional MRI with and without contrast is superior to CT/CTA or DSA for the detection of non-aneurysmal causes of third nerve palsy. Unfortunately, for logistical and insurance purposes, it is usually easier and more economical to obtain an MRI and MRA than a MRI and CTA because MRI and CTA require two separate machines. The other option is to obtain the CTA and inspect the MIP images for structural lesions along the course of the oculomotor cranial nerve. This however, may not display subtle oculomotor cranial nerve or cavernous sinus enhancement, seen in some inflammatory disorders, which MRI would detect. Finally, DSA is still the definitive modality when evaluating for intracranial aneurysms however, many institutions are using only CTA in the initial diagnosis and treatment of intracranial aneurysms.

The Canadian and United Kingdom Perspective

Introduction

Oculomotor cranial nerve paresis caused by aneurysmal compression is a neuro-ophthalmic emergency given the potential morbidity and mortality that is associated with ruptured cerebral aneurysms [32]. Various authors have estimated that aneurysmal compression accounts for only 6–16% of all cases of oculomotor nerve paresis in adults [33]. We also know that unruptured aneurysms are at the greatest risk of rupture within the first 48 h [34]. The physician must therefore make a timely diagnosis of an aneurysm and must weigh the diagnostic accuracy of each imaging modality and know how quickly each of the various imaging modalities can be arranged in the region that the physician practices.

The History

Regardless of imaging modality that is available or used to establish the diagnosis, it is helpful to optimize the contributions that the history and examination make to the pretest probability. We agree with Dr. Vaphiades that the presence or absence of pain is not a reliable discriminating feature and simply reflects that fact that branches of the trigeminal nerve variably travel with the oculomotor nerve as it courses through the cavernous sinus and orbit. We would also stress the importance of examining the abducens and trochlear nerves by asking the patient to direct their gaze to the same side as the oculomotor paresis (to test the abducens) and then downward observing for incyclotorsion (to test the trochlear). If either the abducens and/or trochlear nerves are paretic or if the trigeminal nerve is involved, the focus of diagnostic imaging shifts to optimal viewing of the cavernous sinus. Another point of emphasis is the examination of the pupil. Recognizing synkinesis of pupil suggests chronic compression of the oculomotor nerve. While this does not rule out aneurysm, it raises the possibility of other slowly compressive entities. Mild degrees of anisocoria may leave the investigator with some uncertainty. In these cases, it is helpful to also assess the robustness of constriction to very bright light (use contralateral eye as a control to compare the efferent pathways).

Choice of Imaging Modality

The availability of MRI scanners in Canadian provinces varies considerably but averages 9/million. There are very few units that offer 3 T MRA. There are still many regions of Canada where patients have no local access to MRI and must travel great distances if an MRI is needed [35]. The availability of CT scanners averages 15/million and there are far greater number of these units that have CTA capability [35].

In study of access to diagnostic imaging performed in 2016, Canadian patients could expect to wait time of 3.7 weeks for a CT scan and 11.1 weeks for an MRI scan [36]. These are average wait times and do not factor triaging for cases that potentially threaten life or limb. In such cases the current Canadian Radiologists Association recommendations are that MRI scans should be performed with 24 h [37]. However, according to a study performed in 2009, only 42% of Canadian Centers had documented guidelines for prioritization. Target timelines for each prioritization category varied widely and 16% of centers were not able to meet their target timelines for any prioritization category [38]. Access to diagnostic imaging in the UK is similar in that the median time to test from request is considerably shorter for CT scanning than MRI. Currently, for a CT scan in the UK, the median delay between request and imaging is currently 1 day but the median for MRI is currently 22 days. These figures need to be interpreted with an element of caution as they include all requested imaging from the emergency room, inpatient, outpatient and requests from family practitioners for direct access imaging. The median waits for inpatients and emergency room requests are much shorter but access to MRI is still more challenging than access to CT scanning [39]. Given these Canadian realities, CTA is easier to obtain and quick to perform and has become the preferred initial modality for isolated cases of oculomotor paresis where aneurysms must be ruled out. Fortunately, in recent large cohort study, CTA to have a high accuracy for detection of small cerebral aneurysms (smaller than 5 mm) [40]. This study contradicts earlier studies that suggested CTA performed poorly with small aneurysms [41]. MRA remains the preferred initial choice of imaging for non-isolated cases involving the cavernous sinus.

DSA is only available at some urban centers in Canada where established intravascular treatment programs exist. Even where DSA is available, the expertise of the operator, complication rate and interpretation of the study varies with the time of day that the procedure is performed. Nevertheless, DSA remains the gold standard and has been shown to offer information that alters treatment plans in 19% of plans based on CTA and in 30% of plans based on MRA [42].

Aneurysmal compression of the third cranial nerve is treated as an emergent event due to the morbidity and mortality associated with subarachnoid hemorrhage caused by aneurysmal rupture. The greatest risk of rupture occurs early after symptomatic presentation of an aneurysm. Therefore diagnostic imaging and treatment should both follow rapidly after presentation to minimise both morbidity and mortality.

Historically, treatment has involved open craniotomy and clipping of the neck of saccular aneurysms to exclude the aneurysm sac from the intracranial circulation. There is a high rate of surgical success, including resolution of the oculomotor function following open clipping. There is however a significant degree of morbidity (epilepsy, stroke, post-operative infection) and a low level of mortality associated with this technique [43].

The advent of endovascular techniques to manage intracranial aneurysms has resulted in a significant shift in approach to the treatment of both saccular and fusiform aneurysms of the intracranial circulation. Detachable coils of relatively inert metals, such as titanium, have allowed interventional radiologists to pack the aneurysm sac with a “coil ball” thus obliterating the aneurysm sac and thereby excluding the fundus from the circulation. Initially it was felt that the aneurysm sac needed to be completely obliterated to protect against the risk of subsequent rupture. This lead to the concept of supramaximal filling with some interventionists aiming to produce coil balls that were effectively larger than the measured size of the initial aneurysm sac [44]. The net result of this approach was that a larger space-occupying lesion (SOL) was produced and excluding the aneurysm from the circulation did not relieve signs or symptoms due to compression from this SOL.

Coiling technique has now altered significantly with greater emphasis on excluding the fundus from the circulation and less emphasis on complete obliteration of the aneurysm neck. This has resulted in stable exclusion of the aneurysm sac over years with a reduction in the symptoms and signs due to aneurysmal compression of adjacent structures such as the oculomotor nerve [44].

There have also been other advances in endovascular management that have increased the range of saccular aneurysms amenable to endovascular management and lead to the ability to manage some fusiform aneurysms in a similar manner to saccular aneurysms [45]. Stents composed of an expandable wire mesh were first introduced to allow safe coiling of aneurysms where the neck morphology would not allow coils to reside entirely within the aneurysm. The wire mesh placed across the aneurysm neck in the parent artery produced a barrier to coil loops impinging, or extending into, the circulation of the parent artery as well as allowing access for the microcatheter to deploy coils into the aneurysm sac [46]. Further modification of the configuration of the mesh has led to the development of flow diverting stents such as the Pipeline, which are now used as primary treatment for some aneurysms. These stents effectively maintain the laminar flow of the blood column within the parent vessel, preventing egress of circulating blood into the sac of the aneurysm and thus allow thrombosis and shrinkage of the aneurysm sac [47]. This approach results in significant reduction of any space occupying effect from the initial aneurysm sac, while at the same time negating the risk of subarachnoid haemorrhage due to aneurysmal rupture.

As the field of interventional endovascular management is changing so rapidly, the literature on morbidity and symptomatic relief of oculomotor deficits from endovascular management is not reflective of current practice.

The Role of Monitoring Presumed Ischemic Oculomotor Paresis

Role of monitoring pupil size in the patient with highly probable ischemic oculomotor paresis may be appropriate in some cases (e.g. long-standing diabetics presenting over a week after symptom onset). In these cases, physicians could instruct the patient/family member on how to monitor the pupil size. One must always remember that that pupils may become involved up to 10 days after the initial aneurysmal compression of the oculomotor nerve. The initial status of the pupil is not always helpful. As pointed out by Vaphiades and Roberson, ischemic oculomotor paresis may result in anisocoria and, conversely, compressive oculomotor paresis may present with isocoria. Nevertheless, asking patients to monitor their own pupils and report any suspected change in the difference between the right and left pupil diameters adds an additional layer of safety for the rare occurrence when a presumed ischemic paresis is actually an aneurysmal compressive paresis. Given the high sensitivity of CTA, it is still preferable to order a CTA regardless of how probable an ischemic etiology is. Given rarity of this clinical situation and the rarity of the observer being convinced of this constellation of clinical features which lead to the situation, it is probably safest to arrange urgent CTA imaging of all isolated new onset III CN palsies. However, in some instances, it is not possible to get a CTA right away and asking the patient to monitor and report changes in the pupil or degree of ptosis potentially avoids missing the remote possibility of an expanding compressive aneurysm. There is only one setting in which observation without urgent neuroimaging to exclude aneurysmal compression of the third cranial may be an appropriate course of action. That is the situation where the patient has known vasculopathic risk factors, there is complete paralysis of all of the extrinsic ocular muscles supplied by the third nerve (including the levator) but a normally functioning pupil without anisocoria. The rarity of this clinical situation and the rarity of the observer being convinced of this constellation of clinical features lead to the situation that in reality that it is probably safest to arrange urgent CTA imaging of all isolated new onset III CN palsies. Current imaging assessment, in Canada and the UK, is determined by access and risk of non-detection of significant pathology. The net result is that patients that require imaging will undergo CT angiography usually on the same day in the UK and Canada. Given the caveats above, this now means that all patients presenting with new onset oculomotor palsies have an urgent CTA and decisions on further imaging or management are based on the presence or absence of a detected causative aneurysm.

Acknowledgment

This work was supported in part by an unrestricted grant from the Research to Prevent Blindness, Inc., New York, NY.

A.J.S. is funded by an NIHR Clinician Scientist Fellowship (NIHR-CS-011-028).

There are no commercial or financial conflicts of interest and any funding sources by either author.