2025 Volume 19 Issue 1 Article ID: cr.2025-0033
Objective: We report a patient with occlusion of the distal internal carotid artery (ICA), in whom angiography during mechanical thrombectomy revealed a shunt between the ICA and the cavernous sinus.
Case Presentation: A 79-year-old man with bile duct cancer, a liver abscess, septic shock, and atrial fibrillation presented to our hospital with sudden disturbance of consciousness, conjugate eye deviation, and right hemiplegia. A cranial CT revealed a hyperdense middle cerebral artery (MCA) and loss of gray-white matter differentiation, suggesting large vessel occlusion. Endovascular therapy was immediately initiated. Left internal carotid angiography indicated occlusion of the distal ICA at the origin of the ophthalmic artery. Injection of contrast medium at a site just proximal to the ICA occlusion depicted the cavernous sinus and inferior petrosal sinus. We withdrew the aspiration catheter to the petrous segment of the ICA and injected contrast medium again. This time, however, neither the cavernous sinus nor the inferior petrosal sinus was visualized. We deployed a stent retriever at the occlusion site and successfully removed the thrombus. The final angiography showed complete recanalization of the affected arterial segment with no sign of a carotid cavernous fistula. The patient was finally discharged on day 73 after endovascular therapy with a cerebral infarction in the territory of the left MCA.
Conclusion: In the present case, angiographic visualization of the cavernous sinus varied depending on the site of contrast medium injection. It appears that the high pressure of the contrast medium generated in the stump of the ICA opened up microvascular shunts between the normal capillaries of the ICA and the cavernous sinus, leading to visualization of the cavernous sinus. Therefore, it is important to be aware that injection of contrast medium into the blind alley of the ICA near the cavernous sinus could result in early visualization of the cavernous sinus.
Endovascular thrombectomy is well established as a highly effective treatment option for acute ischemic stroke caused by proximal, large vessel occlusions.1) However, manipulation of various devices, such as catheters and microguidewires, can sometimes cause serious vascular complications, including a carotid cavernous fistula (CCF).1–7)
Herein, we present a patient with distal occlusion of the internal carotid artery (ICA) in whom a rare angiographic pattern, which could have been confused with that of CCF, was visualized during endovascular treatment.
A 79-year-old man with a modified Rankin Scale (mRS) score of 0 and atrial fibrillation was admitted to the surgical department of our institute with bile duct cancer, a liver abscess, and septic shock. On admission, we discontinued oral edoxaban, which he was receiving for atrial fibrillation, because he required invasive treatments, including biliary drainage. On day 21, he was found unconscious with a Glasgow Coma Scale score of E1V1M5, conjugate deviation of the eyes toward the left, and right hemiplegia. His score on the National Institutes of Health Stroke Scale was 20 points. Cranial CT imaging showed loss of gray-white differentiation in the left cerebral hemisphere and a hyperdense appearance of the left middle cerebral artery (MCA) (Fig. 1). The Alberta Stroke Program Early CT Score was 6. Considering the clinical course, we strongly suspected occlusion of the left ICA and MCA. We elected to perform mechanical thrombectomy without intravenous thrombolysis because of his advanced age, and immediately initiated the procedure approximately 138 min. after the patient was last known to be well.
An 8-French balloon guiding catheter (Optimo; Tokai Medical Products, Aichi, Japan) was placed in the left ICA via the left femoral artery under local anesthesia. Left internal carotid angiography, performed by manually injecting the contrast medium using an aspiration catheter, revealed occlusion of the ICA just distal to the origin of the ophthalmic artery (Fig. 2A and 2B). We attempted mechanical thrombectomy using both a stent retriever (Solitaire X 6.0 × 40 mm; Medtronic, Irvine, CA, USA) and an aspiration catheter (AXS Vecta 71; Stryker Neurovascular, Fremont, CA, USA). A microcatheter (Phenom 021; Medtronic) and a microguidewire (CHIKAI 14; Asahi Intecc, Aichi, Japan) were guided to the M1 segment of the left MCA to advance the stent retriever. To accurately identify the occlusion site, we manually injected contrast medium simultaneously through the microcatheter and aspiration catheter (Fig. 2C and 2D). At this time, the left cavernous sinus and left inferior petrosal sinus were visualized. We then pulled back the tip of the aspiration catheter to the petrous portion of the ICA and manually injected contrast medium again. This time, however, the left cavernous sinus was not visualized (Fig. 2E). Because of the low likelihood of development of a CCF by that time point, we continued with the endovascular treatment. Finally, we achieved successful revascularization (modified thrombolysis in cerebral infarction 3) after 2 attempts at thrombectomy using a stent retriever alone (Fig. 2F and 2G). The time from puncture to recanalization was 121 min. The final angiography using an automatic power injector showed no evidence of a CCF. MRI obtained the following day showed infarction of the left MCA territory and maintained patency of the left ICA and MCA (Fig. 3A and 3B). The patient was discharged 73 days after the mechanical thrombectomy, with an mRS score of 5.
In the present case, injection of contrast medium into the left ICA just proximal to the occlusion led to visualization of the cavernous sinus and inferior petrosal sinus, which could be confused with a CCF.
A CCF is caused by mechanical microtrauma with denudation of the endothelium and adjacent vessel wall tissue layers.4,7) Predisposing factors for iatrogenic CCF include advanced age, tortuosity of the ICA, atherosclerotic vessel wall changes, cavernous carotid aneurysms, and vessel wall fragility due to underlying connective tissue weakness.2) Mechanical thrombectomy can also cause an indirect CCF through damage to the meningohypophyseal trunk caused by the devices used.8) Accordingly, neurointerventionists need to maintain appropriate vigilance for the development of a CCF as a complication of mechanical thrombectomy. However, based on the findings in the final internal carotid angiogram, we ruled out the development of CCF in our patient reported.
The contrast medium initially injected into the vicinity of the occlusion site in this case probably generated a high pressure of the iodine agent directly on the vascular wall. It should be noted that physiological arteriovenous shunts exist in the cerebrum and dura mater.9–11) Such arteriovenous shunts can open up or dilate in pathological states.10,12,13) In theory, acute occlusion of the ICA and subsequent cerebral ischemia could dilate the intracranial microvascular arteriovenous shunts. Taken together, the most plausible explanation for the early visualization of the cavernous sinus in the present case is that a high pressure from the contrast medium, generated by injection of the contrast medium into the blind alley of the ICA near the cavernous sinus, resulted in the flow of contrast medium from the ICA through the microvascular arteriovenous shunts into the cavernous sinus. To the best of our knowledge, there are no reported cases in which injection of contrast medium at a site just proximal to the occlusion led to visualization of physiological arteriovenous shunts between the occluded ICA and the cavernous sinus.
Most direct CCFs are high-flow lesions. The high pressure inside the cavernous sinus causes expansion of the affected cavernous sinus and engorgement of the draining vessels, such as the superior and inferior orbital veins, pial and cortical veins, sphenoparietal sinus, Sylvian veins, superior and inferior petrosal sinuses, and the pterygoid plexus.14,15) Cortical venous reflux into the superficial middle cerebral vein has been reported to be visible in 66.7% of cases with direct CCFs.16) The venous drainage is multidirectional.14)
However, previously reported cases of direct CCFs with small shunt flows caused by endovascular thrombectomy did not necessarily show typical angiographic findings, such as retrograde venous drainage and the arterial steal phenomenon.5)
Similarly, none of the above-mentioned angiographic findings typical of the CCFs, except for the visualization of the ipsilateral cavernous sinus and the inferior petrosal sinus, were detected in the present case.
Accordingly, it is difficult to determine during emergent endovascular treatment whether newly appearing stagnation of contrast medium in the cavernous sinus indicates the development of an iatrogenic CCF or the opening of physiological arteriovenous shunts.
If stagnation of contrast medium in the cavernous sinus is still visible in the final angiographic image, it is difficult to completely rule out the possibility of arterial dissection and iatrogenic CCF.
The early visualization of the cavernous sinus indicated in the present case with ICA occlusion is also similar to that seen in cases of cavernous sinus dural arteriovenous fistula supplied only by the dural branches of the ICA.
However, an occult cavernous sinus dural arteriovenous fistula may not be visualized on a diagnostic internal carotid angiogram if the shunt flow is very small; therefore, we could not fully exclude the possible presence of an occult cavernous sinus dural arteriovenous fistula in the present case.
In cases of acute ischemic stroke, rapid recanalization of the occluded vessel is associated with improved outcomes.17) Accordingly, accurate and quick interpretation of unexpected angiographic images during endovascular thrombectomy is crucial. We should be aware of the possibility of visualization of the cavernous sinus through physiological arteriovenous shunts between the ICA and the cavernous sinus.
In this report, we describe a patient with acute occlusion of the distal ICA in whom a physiological shunt between the ICA and the cavernous sinus was visualized during the endovascular treatment. It should be borne in mind that injection of contrast medium into the vicinity of an occlusion in the ICA could result in visualization of cavernous sinus owing to the flow of contrast medium from the ICA to the cavernous sinus via physiological arteriovenous shunts.
All authors have no conflict of interest.
