Symptomatic Cavernous Sinus Aneurysms: Management and Outcome After Carotid Occlusion and Selective Cerebral Revascularization

Delayed Infarctions and Cortical or Subcortical Deficits

Follow-up neuroimaging identified a delayed infarction in one of the eight moderate-risk patients and in none of 16 low-risk patients. None of the low-risk patients developed delayed symptomatology suggestive of postocclusion ischemia or infarction. In the moderate-risk group, only the patient with delayed radiographic infarction developed symptoms. This patient was an 80-year-old woman with a symptomatic left cavernous sinus aneurysm. She had an unremarkable clinical BTO, but baseline and quantitative BTO showed decreased perfusion bilaterally with the left ICA distribution being most affected (Fig 2). The patient underwent a left superficial temporal artery-to-middle cerebral artery bypass followed by an endovascular petrous carotid occlusion. The patient had an unremarkable postoperative course, but developed right-sided weakness approximately 6 months after carotid occlusion with bypass. CT confirmed a new left posterior middle cerebral artery infarction corresponding to the patient’s symptoms (Fig 3).

• Download figure
• Open in new tab
• Download powerpoint

Fig 2.

Example of a xenon-enhanced CT CBF analysis for a patient at moderate risk for ischemic infarction after carotid occlusion. Quantitative BTO CBF analysis (BTO R ICA) identified decreased perfusion of the right ICA. The patient then underwent cerebral revascularization followed by carotid occlusion, with restored baseline CBF postoperatively. STA indicates superficial temporal artery. Each column represents a different brain level of CBF 5 min apart.

• Download figure
• Open in new tab
• Download powerpoint

Fig 3.

Postocclusion infarction predicted by quantitative BTO in an 80-year-old woman who presented with left third, fourth, and sixth nerve palsies.

A, Digitally subtracted cerebral angiogram of the ICA (AP view) reveals a left cavernous ICA aneurysm. The patient underwent clinical BTO and did not develop neurologic symptoms.

B, Baseline and C, subsequent quantitative BTO by means of xenon-enhanced CT CBF analysis reveal decreased perfusion in the left ICA distribution, as well as worsening hemispheric asymmetry with temporary BTO. The patient then underwent a left superficial temporal artery-to-middle cerebral artery bypass followed by carotid occlusion.

D and E, Repeat images obtained at 6-month follow-up show a new left posterior middle cerebral artery branch infarction.

For B and C, top row shows correlative CT image level of CBF measurement and bottom row represents xenon-enhanced CT-CBF analysis.

Signs and Symptoms Related to Cavernous Sinus Aneurysm

Of the 26 patients with cavernous sinus aneurysms, 24 (92%) initially presented with diplopia secondary to cranial neuropathies, six (23%) with severe headache, five (19%) with a Horner’s syndrome, three (12%) with trigeminal symptomatology, and three (12%) with “blurred vision” (Table 1). The most common cavernous cranial neuropathy involved the abducens nerve (69%), whereas the trigeminal nerve was least affected. After carotid occlusion, seven patients (27%) described the development of new symptoms, worsening of existing symptoms, or were found to have new cranial neuropathies on examination (Table 2). During this same period, however, 25% of patients noted improvement in their diplopia, 66% were found to have improved visual ability to focus, and 33% noted improvement in headache (Table 1).

At latest follow-up (mean, 15.3 months; range, 1.1–39 months), most of the new signs and symptoms identified in the initial postocclusion period had resolved. Two-thirds of patients had noted improvement in their diplopias, with nearly 90% of all abducen palsies improved (Table 1). All patients presenting with “blurred vision” described improved visual clarity, and 50% of patients presenting with headache noted improvement. However, patients presenting with complete ophthalmoplegia or trigeminal sensory symptoms were less likely to obtain clinically significant improvement in their symptoms. Only two patients had postocclusion signs and symptoms not identified before carotid occlusion. One patient developed complete ophthalmoplegia after presenting with a partial occulomotor paresis, and one patient developed a hypoglossal paresis after carotid occlusion that was not identified on preocclusion testing (Table 2). The ophthalmoplegia, which developed in the early postocclusion period, likely represented ischemic or compressive injuries to the structures within the cavernous sinus due to aneurysm swelling and thrombosis after occlusion. As new microcollateral vessels to the cavernous sinus structures form and the acute thrombotic inflammatory response resolves, these neurologic deficits might improve. The hypoglossal paresis cannot be described by a local mass-effect phenomenom within the cavernous sinus, but rather may reflect an embolic injury to the microvasculature supplying the hypoglossal nerve at the time of carotid occlusion.

Fifteen patients had postocclusion follow-up of more than 1 year’s time. Fourteen (93%) of these 15 patients had improvements in their preocclusion symptoms. One patient, who presented with “blurred vision” and a partial occulomotor paresis, developed complete ophthalmoplegia after carotid occlusion. Her subjective vision improved, but her ophthalmoplegia did not. Five patients had persistent partial occulomotor nerve palsies (all were improved from before carotid occlusion), and three had abducen pareses (all improved from before carotid occlusion). Eight patients were completely asymptomatic, with normal neurologic examinations.

Local Cavernous Sinus Aneurysm Symptom Control after Carotid Occlusion

Treatment indications in our study for patients with cavernous sinus aneurysms were based on the Linskey et al (32) retrospective study of the natural history of cavernous sinus aneurysms. Indications for treatment included symptomatic idiopathic cavernous sinus aneurysms that were defined by progressive ophthalmoplegia or visual loss, severe ipsilateral facial or orbital pain, or demonstrable aneurysm enlargement based on imaging criteria.

Unlike previously published series on the management of cavernous sinus aneurysms, we evaluated patient symptoms before treatment, early after treatment, 1 year after treatment, and at last evaluation. Twenty-seven percent of the patients in our series developed new symptoms or had progression of their preocclusion symptoms in the early postocclusion period. These symptoms often lasted several weeks and generally resolved within 1–3 months. At last follow-up, only two patients had persistent neurologic deficits not identified by preocclusion testing. These temporary symptoms developing in the early postocclusion period likely represent ischemic or compressive injuries to the structures within the cavernous sinus owing to aneurysm swelling and thrombosis after occlusion. With resolution of the acute thrombotic process, neurologic deficits usually improved. In addition, 93% of patients at 1 year after carotid occlusion had either complete resolution or marked improvement of preocclusion symptoms.

Role of Quantitative CBF Analyses in Evaluating Hemodynamic Risk

The goal of an ancillary study to a clinical BTO is to allow for selective bypass surgery by identifying the patient that develops a near exhaustion of hemodynamic reserve with carotid occlusion from the patient who does not. Presumably such an individual could experience infarction with any additional compromise of perfusion pressure or collateral supply, both of which commonly occur during sleep with a reduction of pressure and a potential rise of Pco₂. The paradigm chosen for this report identified patients with a quantitative CBF in whom BTO reduced flow from a norm of 50 down to a range of 20–30 mL/100g/min. The physiologic reliability of this strategy was supported in the report by Witt et al (19) of CBF changes that accompanied BTO in 156 patients who passed clinical BTO. In that report, no patient had a middle cerebral artery flow value below 20 mL/100g/min, although 14 patients developed flow in the 20–30 mL/100g/min range. This strategy has been further validated by the recent work of Marshall et al (33) in which they prospectively evaluated the predictive validity of quantitative CBF measurements during ICA BTO by using intracarotid xenon-133 injection and scalp scintillation recordings. In their study, a CBF of less than 30 mL/100 g/min during BTO was the only variable predictive of stroke after carotid occlusion. They found that the negative and positive predictive values for quantitative CBF findings were far better than those for clinical neurologic examination or sustained-attention tests.

Although many strategies have been used to identify patients at hemodynamic risk, few clinical series have been reported in which such information was used in guiding treatment options. The report by Origitano et al (27) with a greater than 40% associated stroke rate provides an example of the many potential confounders that can complicate the predictive power of BTO testing. Their use of a qualitative CBF method was also problematic because of the inability of such a study to distinguish with confidence the group that drops flow unilaterally (19). Quantitative CBF analyses, however, remove the potential problems seen with qualitative CBF evaluations when low-flow states exist bilaterally, as seen with our patient in Fig 3. With use of qualitative measures, this patient, who later developed ischemic infarction, would not have been identified as having hemodynamically compromised CBF, as hemispheric flow was fairly symmetric bilaterally. Thus, the use of qualitative measures such as perfusion CT, diffusion or perfusion MR imaging, or SPECT HMPAO may not adequately predict ischemic risk in patients with bihemispheric or multiterritorial disease.

Although xenon-enhanced CT has been used extensively in the past, it is currently only available in the United States as an IND experimental status technology, as is PET CBF analysis. It is anticipated that both of these technologies, as well as Xe¹³³ CBF and Xe¹³³ SPECT (which are approved by the U.S. Food and Drug Administration), will become models for quantitative measures of CBF in the future, to which our results can be applied. However, since xenon-enhanced CT has only recently been approved for experimental use in humans in the United States, our results can be and should be applied when using alternative technologies that measure quantitative CBF such as PET quantitative CBF, intraarterial Xe¹³³ CBF, and Xe¹³³ SPECT (34). Technologies such as transcranial Doppler sonography have also been used in the guided management of carotid occlusion, but because transcranial Doppler sonography measures blood velocity it can only indirectly estimate tissue CBF (35). Therefore, our results may not be applicable to its use.

Ischemic Infarction After Carotid Occlusion with Selective Revascularization

In our series, no patient identified as being low risk by clinical BTO plus quantitative BTO developed symptomatic or radiographic evidence of ischemic infarction with a mean clinical follow-up of more than 15 months and radiographic follow-up of more than 10 months. Thus, the combination of clinical plus quantitative BTO in the evaluation of these patients resulted in a 0% false-negative rate. When comparing our results with those of other previously published series, this combination of clinical plus quantitative BTO also appears to be more sensitive in identifying patients who do not require revascularization when performing carotid occlusion than is clinical BTO alone (6, 12, 13, 27–29). Unfortunately, a true- or false-positive rate could not be determined with this protocol as all moderate-risk patients underwent extracranial-intracranial bypass surgery and thus were presumably no longer as severely hemodynamically challenged. Because no previously published series excluded these potential confounders, they could not assess the true risk of infarction secondary to carotid occlusion in patient groups by using BTO and/or other physiologic technologies.

Embolic Infarction After Carotid Occlusion

Thromboembolic infarction is a common cause for neurologic deficit after carotid occlusion and is reported to occur in 5–10% of patients, despite a normal BTO (15). Furthermore, risk for thromboembolic infarction cannot be assessed by using quantitative BTO or other hemodynamic studies. The use of periprocedural anticoagulation and postocclusion antiplatelet therapies have reportedly decreased this risk over the past 10 years, as has the increased utilization of the endovascular technique to perform carotid occlusion (2, 36–38). Yet, despite the incorporation of these modalities, embolic infarction is still reported.

In our series, all patients underwent periprocedural anticoagulation by using a 7000-U intravenous heparin bolus administered during balloon catheter insertion for BTO or carotid occlusion. Daily aspirin (325 mg orally) was then started 24 hours after carotid occlusion with or without revascularization and continued for life. In ddition, careful attention was made to duplicate the technique of carotid occlusion between the BTO and permanent occlusion procedures. Although, in our patients a combination of endovascular detachable balloons, nondetachable balloons, and coils were used in the carotid occlusion technique, permanent occlusions were performed as proximal to the aneurysm neck as possible. Previous literature has suggested that by decreasing the dead space you may decrease the risk of thomboembolism presumably by reducing the likelihood of collateral recanalization of the aneurysm (39). In the few patients in whom the BTO was performed in the cervical ICA, the distal vessel and aneurysm were injected with contrast material to confirm flow stasis. If contrast material washout was seen, the balloon was advanced further and the process repeated until contrast material stasis was confirmed. Then, the clinical and quantitative BTOs were performed. When the patients later underwent permanent carotid occlusion, the balloon(s) and/or coils were placed at the location where the BTO was performed. We believe this was important in minimizing the risk of both ischemic and embolic infarction by ensuring aneurysm flow stasis. If endovascular carotid occlusion were performed more distally than where BTO was studied this could have resulted in occlusion of collateral vessels not occluded during BTO, causing unpredicted infarction. Similarly, more proximal occlusion could have resulted in collateral flow via vessels that were occluded during BTO, resulting in persistent aneurysm filling. As a result of our method of carotid occlusion and BTO testing, no resultant embolic infarctions were seen and all aneurysms were thrombosed at last angiographic evaluation. Thus, we advocate performing test occlusion analyses at the location of the expected permanent occlusion.