Delayed mid-basilar artery stenosis following paediatric acute mechanical thrombectomy: a rare complication from a rare case

Abstract

There is limited literature on the use of mechanical thrombectomy with stent-retrievers in paediatric stroke, especially in the posterior circulation. We report a paediatric case of acute basilar artery occlusion successfully treated by mechanical thrombectomy using stent-retriever but complicated by delayed basilar artery stenosis. This case emphasises prudent selection of thrombectomy device and meticulous long-term follow-up when employed in paediatric stroke.

Background

Our anecdotal experience reinforces the efficacy of mechanical thrombectomy in paediatric basilar artery occlusion, even several hours after onset. However, notwithstanding the life-saving nature of the procedure, our case demonstrates the potential to cause delayed stenosis which mandates diligent long-term radiological and clinical follow-up. Our case also highlights the need to balance the relative advantages and disadvantages of endovascular thrombectomy techniques in choosing the most appropriate procedure for intervention.

Case presentation

A 6-year-old boy with normal developmental milestones and up-to-date immunisation record presented to the emergency department with increasing lethargy and drowsiness. In the preceding 3 weeks, he had complained of occipital headaches and dizziness. On admission he was bradycardic, hypertensive with Glasgow Coma Score (GCS) of 11 (eyes 4; verbal 2; motor 5) and normal pupil size and reactivity to light. He was dysarthric, hypertonic, hyper-reflexic, with grade 3/5 power (Medical Research Council (MRC) scale) and up-going plantar reflexes.

Investigations

Acute thrombotic basilar artery occlusion was suspected at CT of the brain and confirmed on magnetic resonance angiography (MR-A). Areas of diffusion restriction consistent with acute infarction were seen in the ventromedial upper pons and lower midbrain.

Treatment

The patient was immediately transferred to the regional neurosciences centre for emergency mechanical thrombectomy. The procedure was performed on a biplanar system with the patient under general anaesthesia using a 5Fr right transfemoral access. Angiogram through a 5Fr guide catheter placed in the proximal V2 segment of the left vertebral artery showed abrupt cut-off beyond the mid-basilar artery with no antegrade flow into the superior cerebellar artery (SCA) or posterior cerebral artery (PCA) (figure 1A,B). The basilar occlusion was crossed with Prowler Select Plus (Codman Neuro) and Synchro 14 (Stryker neurovascular) microcatheter–guidewire assembly into the P1 segment of left PCA. After poor results of mechanical thrombectomy with 3×20 mm and 4×20 mm Solitaire stent-retrievers, complete recanalisation was achieved using a 6×30 mm Solitaire AB device requiring five passes in total (figure 1C). During each retrieval, manual aspiration was performed with a 60 mL syringe attached to the guide catheter. Completion of angiogram showed Thrombolysis In Cerebral Infarction (TICI) 2B flow (perfusion with incomplete but greater than 50% distal branch filling) with widely patent basilar artery, right PCA and right SCA, and free antegrade flow (figure 1D,E). The distal left PCA and proximal left SCA had filling defects but distal flow beyond these segments was present. The 24-hour postoperative CT brain showed no further infarction, haemorrhage, hydrocephalus or oedema.

Figure 1.

Angiographic imaging pre-mechanical, intra-mechanical and post-mechanical endovascular thrombectomy using the Solitaire device with 1-year and 2-year follow-up. (A and B) Pre-procedure digital subtraction angiography in anterior–posterior and left vertebral artery sagittal working projections showing complete occlusion of the mid-basilar artery (white arrows). (C) Working projection roadmap image after deploying the Solitaire retrievable stent via the left vertebral artery (black arrow). (D and E) Final immediate treatment angiograms in anterior–posterior and left vertebral artery sagittal working projections showing complete re-canalisation of the basilar artery with associated complete neurological recovery. (F) At 13-month follow-up, the patient remained asymptomatic but digital subtraction angiography demonstrated delayed mid-basilar artery stenosis with 50% luminal narrowing (white arrow). (G) At 2-year follow-up, the patient remained asymptomatic but time-of-flight magnetic resonance angiography demonstrated persistent but stable mid-basilar artery stenosis (white arrow).

Post-procedure, therapeutic heparinisation was commenced. He made a gradual neurological recovery to GCS 15, normal tone, MRC grade 5/5 power and normal gait. Haematological and genetic investigations revealed no cause for the thrombosis including a negative thrombophilia screen. He was commenced on life-long aspirin, a 2-year period of warfarinsation, and underwent neuro-rehabilitation.

Outcome and follow-up

At 1-year follow-up, the patient remained without any neurological deficit including full resolution of his pretreatment dysarthria. Cerebral angiography showed an incidental mid-basilar artery stenosis with 50% luminal narrowing (figure 1F). At 2-year follow-up, he remained without any neurological deficit and repeat MR-A showed persistent but stable mid-basilar artery stenosis (figure 1G). He continues to have annual MR-A surveillance to ensure no progression of the mid-basilar artery stenosis. Stenting would be considered if progression of stenosis was detected on MR-A associated with clinical sequelae.

Discussion

Paediatric arterial ischaemic stroke is an important cause of morbidity and mortality, with the posterior circulation involved in 10%–30% of patients.¹ In the absence of recanalisation, mortality from basilar artery stroke approaches 90%, while survivors may have residual deficits including locked-in state.²

In the absence of randomised controlled trials, treatment modalities including intravenous tissue plasminogen activator and mechanical thrombectomy, which have a strong evidence base in adults, are not established in children. Furthermore, clot retrieval devices validated for use in adults have limited use in children.³ In our case, the presence of poor prognostic factors including reduced consciousness and neurological deficits meant the outcome of not intervening was deemed poor. Therefore, mechanical thrombectomy was attempted after obtaining informed consent from the patient’s parents. Endovascular thrombectomy can be performed by stent-retrievers or direct catheter aspiration (figure 2); the former technique was utilised in this case.

Figure 2.

Current endovascular thrombectomy techniques. (A) Stent-retrievers traverse the occluding thrombus, which consequently becomes entrapped within the device and can be removed. (B) Direct catheter aspiration utilises a large catheter navigated to the proximal end of the thrombus and an attached syringe to enable suction and clot removal.

Despite a 24-hour delay between ictus and recanalisation, our patient made a remarkable clinical recovery. This concurs with the emerging evidence of good outcomes from mechanical thrombectomy even with delayed recanalisation.⁴ The better prognosis in children compared with adults has been attributed to the greater neuroplasticity and fewer comorbidities.⁵

Despite excellent patient recovery from mechanical thrombectomy, we are intrigued by the delayed complication of basilar artery stenosis, which may be related to the recanalisation procedure. The data on safety and efficacy of stent-retrievers have been based on short-term (usually 3-month) clinical follow-up, without angiographic correlation or long-term follow-up for sequelae of treatment. Of the two published angiographic studies following stent-retriever mechanical thrombectomy, one reported 50% stenosis in 4 of 117 cases (3.4%) at 3-month follow-up⁶ while the other found delayed 50% stenosis in 3 of 39 patients (7.7%) following use of the Solitaire device at 1-year follow-up.⁷ In both studies, the stenosis was clinically quiescent. These few cases of arterial stenosis following stent-retriever thrombectomy and the anterior circulation were all found in adults. This has not been previously reported in the paediatric population or in the posterior circulation.

Animal studies have shown intimal and medial wall damage following mechanical thrombectomy, especially with wall-contact devices such as stent-retrievers.^8–10 This may explain our finding. In children, the vessel wall may be more fragile and liable to damage. Our patient required five passes with the stent-retriever to achieve satisfactory recanalisation. Hence, intuitively his basilar artery may have sustained a greater degree of intimal damage than in a patient whose recanalisation was completed with a single pass. The delayed stenosis may represent an outcome of vascular repair mechanisms following this intimal damage. Alternatively or concurrently, ischaemic injury of the vessel wall from luminal occlusion itself may contribute towards this phenomenon.¹⁰

The most widely reported advantage of the stent-retriever technique is the higher rate of successful reperfusion as a stand-alone procedure without the help of rescue therapy as compared with aspiration thrombectomy alone.¹¹ Advantages of aspiration thrombectomy as compared with stent-retriever thrombectomy include the lower theoretical risk of vessel damage including perforation, thrombosis or downstream emboli as traversing the occluding thrombus is not necessary, although systematic review and meta-analyses does not demonstrate any difference in these complications between the two techniques.¹¹ Another potential advantage of aspiration is shorter procedural time, with a reported 7–13 min advantage in groin-to-perfusion time versus stent retrieval in recent studies.^11–13 However, there are no significant difference in overall patient outcomes between the two techniques, nor are there any studies reporting significant differences in the rate of occurrence of delayed effects on vessel wall such as stenosis. Further studies are required to evaluate this phenomenon in detail. Similarly, more diligent and long-term imaging follow-up, at least with non-invasive modalities such as MR-A, would enable better understanding of this potential delayed complication.

Learning points.

• Mechanical thrombectomy may be considered in large vessel stroke in carefully selected paediatric patients.

• Delayed arterial stenosis may complicate mechanical thrombectomy.

• Delayed stenosis post-thrombectomy may be due to stent retriever–related wall damage or ischaemic injury to the vessel wall.

• Both stent-retriever and catheter aspiration techniques have their relative advantages and disadvantages, currently without definitive gold-standard endovascular thrombectomy technique.

• Radiological and clinical follow-up extending beyond 2 years may be important to assess pathological evolution and clinical sequelae of delayed arterial stenosis.