Aspiration catheter failure leading to carotid-cavernous fistula during stroke thrombectomy

Abstract

Endovascular thrombectomy in acute ischaemic stroke commonly uses aspiration catheters, either alone or in combination with stent retrievers. The Penumbra Aspiration System (Penumbra, Alameda, California, USA) was first approved by the US Food and Drug Administration in 2007, with low reported device-related complications. We present a case of a previously unreported complication related to malfunction of a Penumbra aspiration catheter during stroke thrombectomy resulting in a carotid-cavernous fistula.

Background

Endovascular thrombectomy (EVT) in acute ischaemic stroke (AIS) is overall a safe procedure, and leads to better functional outcomes in patients compared with medical management alone.¹ The most common periprocedural complications include vessel injury, vasospasm, vessel reocclusion, reperfusion intracranial haemorrhage and access site injuries.^{1 2} Complications caused by device failure are less common, with reports of guidewire breakage, stent retriever detachment and aspiration catheter fracture.³

The Penumbra Aspiration System was approved by the Food and Drug Administration (FDA) in 2007 for EVT in AIS.⁴ Since then, subsequent generations of Penumbra aspiration catheters have been released. The published rate of adverse events related to device malfunction is low, ranging from 0% to 3.5%.^5–7 The JET 7 reperfusion catheter is one of the newest Penumbra aspiration catheters approved for use in revascularisation of patients with AIS from a large vessel occlusion (LVO), with a larger lumen and new catheter body design (‘Xtra Flex’) that is advertised to improve catheter trackabiity and increase thrombus removal. We describe a previously unreported complication that occurred during revascularisation of an LVO in a patient presenting with AIS from an unexpected malfunction of the JET 7 catheter.

Case presentation

A medically complex inpatient in her 60–70s developed acute aphasia and right-sided hemiparesis. National Institute of Health Stroke Scale was 21. Non-contrast CT was negative for acute intracranial haemorrhage, with an ASPECT score of 8. CTA demonstrated a thrombus in the distal left M1 with complete occlusion of the inferior division and partial occlusion of the superior division of the left middle cerebral artery (MCA). CT perfusion with 104 mL core infarct and mismatch volume of 139 mL. Intravenous thrombolytics were administered and the patient was taken for emergent EVT under general anaesthesia.

Diagnostic angiogram redemonstrated the occlusion. A triaxial system consisting of a Neuron Max sheath (Penumbra), JET 7 catheter and Marksman microcatheter (Medtronic, Minneapolis, Minnesota, USA) was advanced into the left internal carotid artery (ICA). The occlusion in the distal M1 and inferior division was successfully crossed with the microcatheter and a Traxcess wire (Terumo, Somerset, New Jersey, USA). A 6 × 40 mm Solitaire X (Medtronic) was deployed across the thrombus, and the JET 7 catheter was advanced without resistance to the face of the thrombus to perform continuous aspiration prior to embolectomy. After 3 min, the stent retriever was retracted completely into the JET 7 and removed. Stent retriever retraction was smooth without any resistance or the application of an inordinate amount of traction. Immediately after removal of the stent retriever, lateral projection hand-injected angiogram through the JET 7, which was maintained in the same position in the M1 segment, using a 10 mL syringe showed modified Thrombolysis in Cerebral Ischaemia (mTICI) 2b reperfusion without any contrast extravasation. A second hand-injected angiogram in the anterior–posterior projection through the JET 7 then showed abnormal opacification of the cavernous sinus (figure 1), consistent with a direct carotid-cavernous fistula (Barrow type A).⁸ A 0.035 inch Bentson wire was advanced through the JET 7 to maintain access to the ICA, but the wire exited from the side of the catheter a few centimetres from the tip, indicating the presence of a catheter fracture (figure 2). The wire was then withdrawn back into the catheter prior to removing the catheter from the patient. The catheter was examined on the back table, which revealed that the distal 2 cm of the catheter was missing. The fractured tip of the catheter was noted to be lodged within the cervical segment of the ICA (figure 3A). Access was then obtained in the contralateral groin, with the plan of snaring the catheter fragment with an 18–30 mm Atrieve snare (Argon Medical, Frisco, Texas, USA). However, by the time access was obtained, the catheter fragment had migrated distally through the fistula into the cavernous sinus (figure 3B). We then successfully snared the catheter fragment with a 4 mm Amplatz Goose Neck microsnare (EV3, Plymouth, Minnesota, USA) from the ipsilateral access (figure 4).

Figure 1.

Abnormal opacification of the cavernous sinus on anterior–posterior projection left ICA angiogram through the JET 7 reperfusion catheter. The tip of the catheter is located in the left M1 segment but contrast is emanating from the catheter fracture site. ICA, internal carotid artery.

Figure 2.

Guidewire exiting the side of the JET 7 reperfusion catheter (arrow) approximately 2 cm from the catheter tip.

Figure 3.

(A) JET 7 catheter fragment lodged in the cervical ICA (arrow). (B) Distal migration of JET 7 catheter fragment (arrow) through the fistula into cavernous sinus with partial occlusion of the fistula. ICA, internal carotid artery.

Figure 4.

Specimen of the fractured distal 2 cm tip of the JET 7 catheter.

Angiogram following catheter fragment removal again showed the carotid-cavernous fistula, and no opacification of the intracranial vessels beyond the site of ICA injury. Contralateral angiography showed good collateral flow to the left anterior cerebral artery and MCA territories via the anterior communicating artery. Reocclusion of the inferior division of the left MCA was noted. Attempts at crossing the left ICA injury with an Echelon-10 microcatheter (Medtronic) and Synchro-14 microwire (Stryker Neurovascular, Fremont, California, USA) were unsuccessful and therefore stent placement across the injury was not feasible. The cavernous sinus was then catheterised and embolised using Micrusframe S and Galaxy G3 coils (Cerenovus, Irvine, California, USA). Once the cavernous sinus was embolised we again attempted to cross the site of injury but were unsuccessful. As there continued to be opacification of the fistula (figure 5A), the left ICA was embolised with additional Galaxy G3 coils and an 8 mm Amplatzer Vascular Plug 4 (Abbott, Abbott Park, Illinois, USA). Completion angiogram showed complete occlusion of the left ICA and no opacification of the carotid-cavernous fistula (figure 5B). Final contralateral angiograms showed preserved collateral perfusion and at this time we observed near complete reperfusion of the left MCA territory with spontaneous reperfusion of the majority of the inferior division (figure 6A). A small residual occlusion in an opercular branch was noted (figure 6B). The patient was admitted to the neurocritical care unit after the procedure.

Figure 5.

(A) Angiogram demonstrating continued opacification of the carotid-cavernous fistula after embolisation of the cavernous sinus. (B) Completion angiogram after embolisation of the left ICA with additional coils and an Amplatzer vascular plug showing no further opacification of the ICA or carotid-cavernous fistula. ICA, internal carotid artery.

Figure 6.

(A) Contralateral angiogram after embolisation of the carotid-cavernous fistula shows collateral perfusion of the left ACA and MCA territories via the anterior communicating artery, and interval spontaneous reperfusion of the majority of the left MCA inferior division (arrow). (B) Lateral view shows a small residual occlusion in an opercular branch (arrow). ACA, anterior cerebral artery; MCA, middle cerebral artery.

Outcome and follow-up

CT 1 day after intervention showed subarachnoid haemorrhage limited to the left sylvian fissure as well as mild gyral oedema of the left cerebral hemisphere (figure 7), which remained stable on a follow-up CT 3 days later. The patient had persistent severe aphasia, hemiparesis, as well as left third nerve palsy and her neurological examination did not improve at the time of her discharge to a skilled nursing facility approximately 1 month after her AIS. No longer term follow-up is available at the time of writing as this was a recent case.

Figure 7.

Non-contrast CT head 1 day after intervention shows subarachnoid haemorrhage limited to the sylvian fissure (arrow).

Discussion

The Penumbra JET 7 reperfusion catheter gained FDA approval in 2019. It has a larger, 0.072 inch lumen compared with its predecessors, and has a tip reinforced with nickel titanium wire.⁹ The catheter is used in conjunction with a vacuum canister, which provides aspiration for thrombus removal. EVT for AIS has evolved over the years, but generally uses a stent retriever, a large bore aspiration catheter, or uses both in combination, often in a coaxial fashion.^{1 10}

Complications from device malfunction are overall rare.^4–7 A search in the FDA’s Manufacturer and User Facility Device Experience (MAUDE) database yielded 18 reports of JET 7 catheter fractures and 15 reports of catheter deformation during use since the date of the catheter’s FDA approval. One of the reports described a breakage in the catheter that occurred during flushing on the back table after it was used for one pass and another described an accordioned distal tip of a JET 7 catheter after two passes and visualised ‘expansion’ of this catheter proximal to the deformed tip during flushing on the back table. Another catheter fracture incidence reported expansion and breakage of the JET 7 catheter during contrast injection through the JET 7 leading to rupture of the patient’s ICA. No prior reports of an iatrogenic carotid-cavernous fistula from fracture of the JET 7 catheter have been reported. In comparison, in reviewing the 111 MAUDE database reports of ACE68 (Penumbra) catheter breakages/deformations during use since its FDA approval in 2016, none were caused by flushing or contrast injection through the catheter.

We do believe that formation of direct carotid-cavernous fistula during this case was directly related to the JET 7 catheter rupture given localisation of contrast egress from the site of the catheter rupture and the fistula. Up to this point in the procedure, there were no indications of possible injury to the carotid artery. As to exact mechanism we hypothesise that the initial arterial injury had been caused by the damaged catheter structure rather than simple contrast extravasation from the catheter rupture site.

Potential reasons for this unfortunate complication may have been related to manufacturing defects, improper back table preparation, or intraprocedural device damage. Of note, a few weeks after this case, Penumbra issued a Notification to Healthcare Providers letter (see online supplemental material) communicating a ‘labelling update’ in the Instructions for Use for the JET 7 catheter, cautioning users to ‘not inject contrast media through the Penumbra JET 7 Reperfusion Catheter with Xtra Flex technology using a syringe or automated high-pressure contrast injection equipment’, which ‘may cause the distal tip of the catheter to expand or rupture’. The warning not to inject contrast with a syringe is unique to the JET 7 and does not exist for other Penumbra reperfusion catheters. This notification proceeds to caution that ‘The Penumbra JET 7 Reperfusion Catheter with Xtra Flex technology has not been tested for compatibility with other manufacturers’ revascularisation devices. The safety and effectiveness of combined use is unknown and could result in damage to the Penumbra JET 7 reperfusion Catheter with Xtra Flex technology.’

Based on the new labelling, interventionalists may want to reconsider the use of the JET 7 catheter, given the availability of other devices.

Learning points.

• Complications from device malfunction can occur and can lead to devastating consequences during endovascular neurointerventions. In the current case, the direct carotid-cavernous fistula complication that occurred during the intervention necessitated endovascular sacrifice of the affected internal carotid artery, which could have potentially affected the patient’s outcome although the patient already had a relatively large core infarct on preprocedural CT perfusion.

• The recent manufacturer warning (‘Notification to Healthcare Providers’ letter dated 27 July 2020) against injecting contrast media through the Penumbra JET 7 reperfusion catheter using a syringe is a new warning for their aspiration catheters. The warning against injection of contrast media with automated contrast injection equipment exists for previous generations of Penumbra aspiration catheters but the warning against contrast injection through the catheter with a syringe is unique for the JET 7.

• Carotid-cavernous fistula is an unusual complication that can occur during endovascular intervention for acute ischaemic stroke. The ability to recognise and treat this complication is essential for all physicians who treat acute ischaemic stroke.