Endovascular treatment of stroke in children under 2 years with heart failure and ventricular assist device

Abstract

Introduction

Strokes in children are characterised by a high mortality rate while, at the same time, the low number of cases makes it difficult to gain practical experience. As heart disease is the most common risk factor, and as more and more cardiological interventions are being carried out, an increase in the incidence of paediatric stroke is expected. In some cases a transplant is required. While waiting for a donor, the use of ventricular assist devices may be necessary. These present with a high rate of neurological complications. We present two cases of children under 2 years of age awaiting heart transplantation supported by ventricular assist devices who had a stroke which was treated by endovascular techniques.

Case 1: A 16-month-old boy with restrictive cardiomyopathy who was listed for a cardiac transplant. At 20 months he required an implantation of an external biventricular support device (Berlin Heart) and had a left hemisphere stroke at 23 months. An intra-arterial approach was used and produced good clinical results. One month later, a heart transplant was performed successfully.

Case 2: An 18-month-old girl with non-compacted dilated cardiomyopathy included in the cardiac transplant programme and in need of a Levitronix Centrimag ventricular assist device presented with an acute left hemisphere stroke at 23 months. An intra-arterial procedure was carried out leading to positive clinical results except for residual right hypertonia. Seven months later she received her transplant.

Conclusion

As a result of the difficulty in performing arterial puncture, the small vessel calibre and the limitation in the use of iodated contrast, there are certain limitations to endovascular treatment of strokes in children that can lead to complications. A multidisciplinary approach to managing such cases would be helpful.

Introduction

Stroke of any aetiology in paediatric patients has an annual incidence of 2.3 to 13 per 100,000 children depending on series. The rate of arterial ischaemic stroke (AIS) is lower (1.2 to 7.9 per 100,000).^1,2 Although the majority of children who have a stroke survive, it carries a significant incidence of mortality.

Children have their own risk factors,³ of which heart disease is the most common one.⁴ Some types of heart disease can only be treated by transplant. While waiting, the use of ventricular assist devices may be necessary. These devices present with a high rate of neurological complications (up to 29%), being one of the main causes of death.⁵

Clinical presentation can be inconclusive, which can lead to a delayed diagnosis.³ Until now the treatment of stroke in paediatric patients has largely been based on supportive therapy. As a result of a lack of grade I evidence, intravenous or intra-arterial fibrinolytic treatment has not yet been included in the guidelines of the American Heart Association (AHA).⁶ The most common approach to the treatment of stroke in the general population has been through intravenous recombinant tissue-type plasminogen activator (rtPA) but with the publication of the Mr. Clean study, ESCAPE and EXTENDIA^7–9 in 2015, it was demonstrated that the treatment of stroke in adults through endovascular techniques is a safe and effective procedure. Thus it is starting to be considered an option for children as well. This fact, together with the expected increase in cases of paediatric strokes which come as a result of advances made in cardiology,¹⁰ means that the treatment of stroke in children is becoming more frequent. In this article, we present two cases of children awaiting a heart transplant supported by a ventricular assist device, who presented with an acute stroke which was treated by endovascular techniques.

Case 1

A 23-month-old boy with hypoxic-ischaemic encephalopathy at birth with good brain motor potential and normal psychomotor development. He had a personal history of restrictive cardiomyopathy and was included in a cardiac transplant programme when he was 16 months old. He also required the implantation of an external Berlin Heart biventricular support device. In order to prevent embolic events, double antiplatelet and anticoagulant treatment was administered. When he was 23 month old presented with disconnection and right hemiparesis. A computerised tomography (CT) scan showed a hyperdense left middle cerebral artery (MCA) (Figure 1(a)), as well as a chronic right parietotemporal infarction. His blood analysis showed: red cells 4.16 × 10⁶µ/L; haemoglobin 11.4 g/gL; activated partial thromboplastin time (APTT) 93 seconds and international normalised ratio (INR) 1.08.

Figure 1.

(a) The cerebral scan showed a hyperdense left middle cerebral artery (MCA) (arrow). (b) The angiography of the left carotid artery showed an occlusion of the M1 segment of the left MCA (arrow). (c) The control run of the right carotid artery showed irrigation of the left hemisphere through the anterior communicant artery.

Intravenous thrombolytic treatment was contraindicated due to double antiplatelet and anticoagulant treatment at full dose with heparin, so an intra-arterial thrombectomy was performed. Although the patient was 23 months old, he was in the third percentile of the weight curve (10 kg). Under general anaesthesia, the right femoral artery was punctured and an 11 cm long 4F sheath (Cordis, Ireland) was placed. A 4F vertebral Radiofocus catheter (Glidecath of Terumo, Belgium) was used to confirm the occlusion of the M1 segment of the left MCA (Figure 1(b)). The artery was recanalised by mechanical thrombectomy with a stentriever using the 4F vertebral catheter as a tutor, positioning it in the petrous segment of the carotid artery. A 3 mm × 20 mm Trevo XP Pro Vue device (Stryker, The Netherlands) was used, with a straight Rapid Transit microcatheter (Codman Neurovascular, UK), opening the artery in one pass. The whole system was removed (tutor catheter, microcatheter and stentriever) simultaneously. In the control run, an iatrogenic dissection of the left internal carotid artery was detected, which had no clinical significance, however, as the left hemisphere was irrigated through the anterior communicant artery (Figure 1(c)).

One month later, his heart transplant was carried out successfully. Neurologically, the only long-term sequel that came as a result of the stroke was a right upper limb spasticity.

Case 2

A 23-month-old girl with a personal history of non-compacted dilated cardiomyopathy that had progressed unfavourably. She was included in the cardiac transplant programme, needing a Levitronix Centrimag ventricular assist device. As in the previous case, she was on double antiplatelets and anticoagulants. When she was 23 months old she presented with irritability and an acute right hemiparesis suggestive of occlusion of the left internal carotid artery, which was confirmed on the CT (Figure 2(a)) and later on the angiography (Figure 2(b)). Her blood analysis showed: red cells 3.41 × 10⁶ µ/L; haemoglobin 9.2 g/gL; APTT 134 seconds and INR 1.2. As in case 1, intravenous fibrinolytic treatment was contraindicated due to double antiplatelet and anticoagulant treatment with heparin at full dose, so intra-arterial treatment was planned. Although she was 23 months old, she was in approximately the third percentile in the weight curve (9.3 kg).

Figure 2.

(a) The scan at diagnosis showed hyperdense left terminal carotid artery and middle cerebral artery (arrow) in relation to occlusion of those vessels. (b) The initial angiogrphy showed an occlusion of the terminal left internal carotid artery. (c) The control run after thrombectomy showed a recanalised left internal carotid artery.

The same technique was used as the one described for the previous case. The vessel was opened in one pass (Figure 2(c)). A disseminated embolus in segment A2 of the left anterior cerebral artery was observed in the angiographic control series. The CT scan at 24 hours showed an established infarct of the territory of the anterior division of the MCA and of the left anterior cerebral artery. The neurological examination showed right hypertonia. When she was 30 months old she received her transplant with a good immediate postoperative result. Three weeks after the transplant, a right ventricular dysfunction was detected, so at the time of writing this article she was still in the intensive care unit.

Discussion

Stroke of any aetiology is not uncommon in children and its prognosis is better than in adults. Despite this, 20% experience long-term neurological damage^4,11 so the social impact is significant.⁴ When it comes to paediatric AIS the first problem facing the medical community is the low number of cases, which makes it difficult to establish protocols. There are many causes in paediatric ischaemic stroke, which is quite different from adult stroke. This is why there is no guideline for treatment in children. In fact, the Thrombolysis in Pediatric Stroke Trial (TIPS) was terminated prematurely because of a lack of recruitment.¹² In 2017, 98 patients received an endovascular treatment at our facility, only two of which (2.04%) were children.

An increase in the incidence of this pathology is expected, because heart disease is the most frequent risk factor and more and more cardiological interventions are being carried out.¹⁰ For children who require ventricular assist devices the likelihood of neurological events increases to 29% and of a second stroke to 25%.⁵ In patients awaiting heart transplant, the treatment of stroke takes on special importance because a major stroke can lead to the removal of a patient from the transplant list.

Delay in diagnosis is common.¹³ This is one of the reasons why endovascular treatment may be the only option because it allows the therapeutic window to be expanded. Although we have no data for extended therapeutic windows in children, it would make sense given the good collateral circulation in the paediatric brain. As is known in adults, the presence of collateral vessels offers protection against acute ischaemic injury, and is associated with improved clinical outcome and reduced infarct volume following thrombolysis.^14,15

Although in the cases we present, there was no delay in the diagnosis, the time to needle could have been shorter if there had been greater coordination among professionals. This pathology should be addressed in a multidisciplinary and coordinated manner. When treating children, most of the time before recanalisation is spent attempting to puncture the femoral artery. Ultrasound assisted femoral puncture should be performed in all cases and in especially difficult cases it would be useful to include paediatric interventional radiologists or vascular surgeons (in centres without paediatric interventional radiologist expertise) in the isolation of the arterial route. Traditionally, the treatment of strokes in children has been based on medical support. Only in highly selected patients is intravenous fibrinolytic therapy with rtPA employed because its use in children is associated with a higher rate of complications.¹⁶ Likewise, the endovascular approach is not performed systematically owing to lack of grade I evidence in the AHA guidelines.⁶ The 2018 AHA guidelines do not introduce any updates regarding treatment in paediatric patients.⁶ The 2013 and 2015 guidelines suggested that endovascular treatment of large vessel occlusions in patients under 18 years is reasonable in selected cases when arteriography begins within 6 hours from the onset (class IIb, level of evidence C).¹¹ A review published in 2016 showed good clinical progress in 85% of children treated endovascularly for acute stroke.¹⁷ As a result of anticoagulation an intravenous approach was contraindicated in our cases, so endovascular treatment was performed because both patients presented with an occlusion of a large vessel and could be treated within the first 6 hours. In older children, the greatest limitation lies in the restriction in the use of contrast.¹⁸ If necessary, series can be performed with diluted contrast and by limiting the number of runs. In the cases we have reported a hyperdense MCA was evidenced on the CT scan, so it was assumed that it was occluded and no CT angiography was performed. As a result, more contrast was available during the intra-arterial procedure. In those cases in which the CT scan allows a clear diagnosis of acute occlusion of a vessel, the vascular study by CT could be avoided. In those centres where magnetic resonance is available, it is also possible to carry out a time of flight sequence in order to explore the vascular system.

In children under 2 years, we must add the size limitation of the introducers and catheters. There is also an inability to use balloon guide catheters in children due to size. Now the avaliable stentriever devices (even the 3 mm wide ones) are designed to be used through a 5F or wider catheter and sheaths. There is not a specific device for paediatric use. A 4F catheter and sheath is not tested for this purpose, even though they have been used in the cases we present, due to the arterial size limitation. This fact involves a high risk of embolus squeezing and migration as could be seen in case 2, in which the anterior cerebral artery was occluded by a migrated emboli. Manual compression of the left internal carotid artery while removing the stentriever would have been a way of avoiding distal emboli migration in case 2. Another alternative would be to use an introducer of a greater calibre (5F), which allows the use of the usual tools. In the absence of a paediatric interventional radiologist this could lead to femoral artery damage and limb ischaemia, which might have to be treated by a vascular surgeon.

Endovascular treatment in paediatric stroke is an alternative but it carries some difficulties that are necessary to bear in mind: the use of iodated contrast is more restricted than in adults, femoral puncture can be difficult and extended, and the use of smaller calibre catheters are limitations that can lead to complications. The correct selection of the patient as well as a multidisciplinary and coordinated management approach are necessary to minimise the risk of complications related to the technique.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.