MRI SPACE sequence confirmation of occluded MCA M2 dissection stump masquerading as a ruptured MCA aneurysm

Abstract

Intracranial vascular pathologies often have overlapping clinical presentations. Dissected vessel occlusions and bifurcation aneurysms can appear similar on pretherapeutic imaging. The medical management of these two entities is drastically different. The patient is a 51-year-old man who presented with severe, sudden-onset headache. Initial presentation was consistent with a ruptured middle cerebral artery (MCA) aneurysm and surgical clipping was recommended. However, further review of radiographic findings could not definitively differentiate an aneurysmal origin of the symptoms as opposed to intracranial dissection followed by occlusion of the M2 branch of the MCA. MRI sampling perfection with application optimised contrasts using different flip angle evolution (SPACE) was performed and showed thin flow signalling distal to the dissected vessel stump confirming the diagnosis. Accurate diagnosis is a crucial step in directing treatment for intracranial vascular lesions. MRI SPACE is a simple tool in the diagnostic armamentarium to adequately direct treatment and avoid the potential for unnecessary interventions.

Background

Important intracranial vascular pathologies such as aneurysms, arterial dissections, vascular loops and occluded vessels may infrequently have overlapping clinical presentations and findings on imaging.^1–8 Herein, we present a case of a dissection-related middle cerebral artery (MCA) branch occlusion mimicking an aneurysmal subarachnoid haemorrhage (aSAH) both clinically and radiographically. Collaborative discussion between the initially treating endovascular neurologist and the subsequently consulted endovascular neurosurgeon resulted in a difference in opinion as to the aetiology of the subarachnoid haemorrhage (SAH). Additional guidance from a neuroradiologist was sought who recommended advanced MRI. The goal was to potentially avoid the need for unnecessary surgical exploration by instead confirming the diagnosis with non-invasive imaging.⁸

Case presentation

History

A 51-year-old man with a history of hypertension presented to an outside emergency department after experiencing a history of sudden-onset, severe, persistent headache over the previous 12 hours associated with acute-onset dysarthria and diffuse paraesthesia of his left hand and arm. The patient reported no nausea, vomiting, nuchal rigidity, photophobia or motor deficits at that time. CT imaging at the referring facility demonstrated hyperdensity in the right Sylvian fissure consistent with acute SAH (figure 1), and the patient was emergently transported to our facility for further evaluation and management. On arrival, the patient was awake and oriented to person, place and time, complaining of an intermittent headache that was relieved after receiving acetaminophen and fentanyl. Prior surgical history was non-contributory.

Figure 1.

Subarachnoid haemorrhage on non-contrasted CT of the head.

Physical examination

On admission, the patient was following commands and was appropriately alert and oriented. Mild left facial droop as well as persistent dysarthria were noted. All other cranial nerves were intact. Motor examination revealed 5/5 strength in the right upper and lower extremities, 4+/5 strength in the left lower extremity and 4/5 strength in the left upper extremity. A pronator drift on the left was present. Sensation to light touch was intact in all extremities. Hoffman sign was negative bilaterally, and deep tendon reflexes were 2+ throughout.

Investigations

Imaging

A repeated head CT without contrast revealed a modified Fisher grade 1 SAH within the right Sylvian fissure extending along the right frontal cortical sulci. Cerebral angiography was initially negative; however, subsequent repeat angiography 1 week later revealed a 3 mm conical-shaped projection from the posterior-superior aspect of the distal right M1 segment of the MCA (figure 2A). The endovascular neurologist diagnosed this as a ruptured right MCA aneurysm that was not amenable to coil embolisation, and therefore, the consultation of a dual-trained neurosurgeon for surgical clipping was obtained. On the neurosurgeon’s review of the patient’s angiography prior to surgery, subtle retrograde filling of a distal MCA branch of the superior division by anterior cerebral artery (ACA) and posterior cerebral artery (PCA) pial collaterals was noticed. This finding combined with the conical shape of the suspected aneurysm suggested an alternate diagnosis of a cerebral arterial dissection with a secondary branch occlusion resulting in an arterial stump mimicking a right MCA aneurysm. Initial MRI revealed focal regions of infarction on diffusion-weighted imaging. Subsequent advanced MRI was performed. MR perfusion demonstrated relative decreased perfusion in the right frontal lobe. T2 sampling perfection with application optimised contrasts using different flip angle evolution (SPACE) sequence was used (TR/TE=1000/136 mm). T2 SPACE is a fast turbo spin echo technique with special modifications optimising it for isotropic 3D imaging.⁹ This sequence showed thin signalling distal to the suspected aneurysmal stump, revealing an occluded vessel with some patency rather than a ruptured cerebral aneurysm (figures 2B–D and 3).

Figure 2.

Comparison of the cerebral angiogram and the MRI sampling perfection with application optimised contrasts using different flip angle evolution (SPACE) sequence. (A) Anteroposterior cerebral angiogram showing a 3 mm conical-shaped projection from the posterior-superior aspect of the distal right M1 segment of the MCA. (B–D) Coronal MRI SPACE sequence showing thin signalling distal to the suspected aneurysmal stump, revealing an occluded vessel with some patency rather than an aneurysm.

Figure 4.

Follow-up cerebral angiography of the right middle cerebral artery (MCA). Follow-up cerebral angiogram showing (A) recanalisation of the MCA branch. (B) The recanalised MCA branch was subsequently coiled to prevent any further complications.

Figure 3.

Comparison of the CT angiogram and MRI sampling perfection with application optimised contrasts using different flip angle evolution (SPACE) sequence. (A–B) CT angiogram and MRI SPACE showing the right middle cerebral artery (MCA) vessel. (C) CT angiogram showing no distal flow in the MCA vessel. (D) MRI SPACE showing small distal MCA vessel that does not correlate with CT angiography.

Treatment

Patient course

On admission to the neurointensive care unit, an SAH treatment protocol was initiated that included systolic blood pressure control <140 mm Hg, daily transcranial Doppler ultrasound, nimodipine, euvolemia and frequent neurological checks. Pain was well controlled with fentanyl intravenous injection. Laboratory findings remained stable throughout the course of treatment. After advanced imaging revealed an occluded vessel, surgical clipping was no longer considered, and the patient was treated conservatively. He was stabilised in the intensive care unit, where his neurological examination improved to near-baseline status prior to discharge.

A follow-up angiogram performed approximately 2 months after the SAH demonstrated a near-complete revascularisation with irregular morphology of the previously occluded vessel (figure 4A). Various treatment options were considered given prior reports of therapeutic success with intracranial stents.¹⁰ However, given the unpredictable natural history of cerebral dissection, revascularisation of this vessel was considered to possibly increase the risk of a recurrent haemorrhage event and the patient was scheduled for elective coil occlusion of the proximal aspect of the previously dissected/occluded MCA branch. Ishihara et al previously demonstrated the safety of coil embolisation of dissection-related posterior inferior cerebellar artery (PICA) aneurysms.¹¹ This procedure was performed approximately 3 months after the initial SAH without complications, and without any change in neurological function. A follow-up angiogram 3 months later demonstrated continued occlusion by the coil mass (figure 4B). The vascular territory of this artery had adequate collateral flow without capillary phase parenchymal loss of any contrast staining. The patient’s final outcome was resolution of all symptoms associated with the initial haemorrhage and infarction.

Discussion

There have been few reported cases of cerebral vascular occlusion mimicking aneurysm, and a majority of those cases involved vessels of the posterior circulation.^1–4 Four of the five reported cases related that MCA branch occlusion mimicking unruptured aneurysms was not recognised until intraoperative inspection,^{5 6 8} while in the fifth case, surgery was avoided because delayed angiography repeated prior to the scheduled surgery revealed recanalisation of the occluded vessel.⁷ This puts an emphasis on the importance of distinguishing between these two diagnoses, as the treatments differ significantly.¹ The advanced imaging obtained in our case prevented the need for unnecessary craniotomy.

In the present case, the patient’s initial symptoms, the natural variation of cerebral vasculature and the limitations of angiographic studies all contributed to the initial interpretation of the vessel stump as an aneurysm. In this case, facial weakness with dysarthria, arm and hand paraesthesias and left-sided weakness are certainly possible after SAH caused by a ruptured aneurysm. However, these focal neurological deficits are much more likely in moderate-grade to high-grade aSAH patients and less likely in an SAH patient who has a Glasgow coma scale score of 15 and World Federation of Neurological Surgeons SAH scale score of 1. The typical clinical presentation of low-grade aSAH does not include focal neurological deficits and as such raised concern in this case by the neurosurgeon. Likewise, there were subtle angiographic signs that raised the possibility of vessel occlusion rather than an aneurysm. Specifically, the retrograde filling of the M3 segments from the ACA and PCA pial collaterals on angiography suggested a more proximal occlusion or stenosis. The location of the suspected aneurysm in our patient raised suspicion for an occluded vessel, as the majority of MCA aneurysms occur at the bifurcation.¹² These signs and symptoms led to a higher index of suspicion for an alternate diagnosis and collaborative discussion resulted in a plan for further advanced radiographic diagnostic modalities.

On recommendation by the neuroradiologist, we decided to use the MRI SPACE (a 3D fast spin echo black blood sequence; Siemens) sequence to further explore this vascular pathology. Sequences similar to the Siemens SPACE sequence are known to differentiate between intracranial intraluminal space, vessel wall and perivascular tissue.¹³ While other similar high-resolution isotropic sequences could have been used (ie, Constructive Interference in Stead State), SPACE sequence was chosen based on operator familiarity, as well as the relative robustness in discerning cerebrospinal fluid (CSF) signal from any other signal.

Our decision to use the MRI SPACE technique resulted in confirmation that the lesion was the stump of an occluded artery and not a ruptured aneurysm. Ultimately, this resulted in our avoidance of an unnecessary operation, whereas, in comparison, four of five previous reports of this identical problem resulted in an unnecessary craniotomy.

A contributing factor to the difficulty in differentiating an aneurysm from an occluded branch is the normal anatomical variation of the MCA branches. Nearly all the case reports of MCA occlusions mimicking aneurysms occurred at or near the division of the M2 segment.^5–8 Though M2 division most commonly bifurcates into a superior and inferior division, this variation only composes 78% of cases, while 12% trifurcate into a superior, middle and inferior division, and 10% form smaller branches. Park et al⁶ described two patients whose angiographic findings suggested the common M2 bifurcation variant with aneurysms at the branching point, though subsequent studies revealed that these patients had trifurcations with atresia of the third vessels. Certainly, less common than an M2 bifurcation, the visualisation of a trifurcation with an aneurysm at the branch point does not eliminate the possibility that the suspected aneurysm may in actuality be an occluded fourth-branch vessel.⁷ As exhibited in our patient, the converse can be true: an atretic vessel at an M2 bifurcation may still be misinterpreted as an aneurysm and may lead to either an earlier branching artery or later M3 bifurcation to be misinterpreted as a more proximal or distal division of M2, respectively. Our patient had an occlusion of the superior division of the right MCA, which led to the branching point of the anterior temporal artery from M1 to appear as the M2 segment bifurcation, giving the impression that the occluded stump of the superior division was an aneurysm arising from the inferior division of the M2 bifurcation. The suspected aneurysm, however, was located at a straight segment, supporting the diagnosis of an occluded stump. The likelihood of a true aneurysmal bifurcation is much more likely with an inclination angle greater than 10%, but a suspected aneurysm shown on angiography at the M2 bifurcation with a minimum inclination angle of 10% should still allow suspicion for an occluded third branch, as discussed above. Our patient’s suspected aneurysm also had a conical shape, which is somewhat atypical for an aneurysm, though a globoid or saccular appearance does not rule out an occluded vessel.⁶

Just as vessel occlusions and aneurysms may have very similar angiographic findings, they may also have very similar clinical presentations. Our patient’s most concerning symptom was the acute onset of a severe headache, which prompted suspicion for a ruptured aneurysm in the context of SAH seen on CT. Of the cases of MCA branch occlusions mimicking aneurysms that were reviewed, none were associated with SAH as in our patient, which made an aneurysmal source of bleeding much more likely when the vessel stump was initially found on angiography. Furthermore, while localised clinical signs of ischaemia without SAH may also occur with large aneurysms secondary to a migratory thrombus, our aneurysm mimic was small. Thus, the additional focal ischaemic symptoms identified in our patient contrasted with a typical small aneurysmal rupture presentation without a coexisting intraparenchymal haemorrhage. This prompted further diagnostic exploration using the MRI SPACE sequence. In this case, with the discovery of the continuation of a narrowed vessel using MRI SPACE, the combination of both localised ischaemic findings and SAH more logically suggests intracranial arterial dissection and secondary arterial thrombosis, which may account for approximately 3% of acute SAHs. The images obtained using the SPACE sequence were successful in excluding the presence of CSF (high signal intensity on T2 SPACE sequence) around the conical abnormality, as would be expected with an aneurysm. The hypointense signal seen distal to the conical abnormality, given its shape, location and signal characteristics, was most consistent with a blood vessel, confirming the expected diagnosis of a dissection stump and sufficiently excluding the presence of an aneurysm.

Occlusions of the cerebral vasculature mimicking an aneurysm can drastically influence treatment decisions for patients. Occluded or severely narrowed vessels and small aneurysms may have very similar angiographic findings and clinical presentations, emphasising that both occlusion and aneurysm should be included in the differential diagnosis when evaluating a patient with a suspected aneurysm seen on angiography. To the uninformed clinician, these similar presentations may result in unnecessary surgical clipping or endovascular coiling. Careful examination of the clinical history, location and shape of the suspected aneurysm, and the appropriate use and careful interpretation of advanced diagnostic tools may allow differentiation between these two aetiologies and avoidance of unnecessary surgical intervention and the associated risks.

Learning points.

• Cerebral arterial dissections/occlusions can masquerade as aneurysms.

• MRI sampling perfection with application optimised contrasts using different flip angle evolution (SPACE) sequences can be useful to identify vessels when they are not patent.

• Recanalisation of vessels following occlusion are better embolised to prevent further embolic complications.