Article Text
Abstract
Background Among the technical challenges involved in safely performing neuroendovascular procedures in infants are limitations on the size of femoral arterial access catheters. Restricting groin access to 4 F and 5 F systems constrains the range of procedures that can potentially be performed.
Methods and Results We present here a consecutive series of transfemoral angiographic procedures in children with no permanent and one transient case of groin morbidity related to femoral access. We illustrate two strategies for using 4 F and 5 F systems for interventions in a manner that enables such techniques as balloon assisted coil embolization and the deployment of triaxial catheter systems to be used.
Conclusions Extremely low groin morbidity is achievable using rigorous parameters around femoral catheterization in children. With creative approaches to working within the limitations of the access vessel size, complex endovascular neurointerventions are possible, even in the youngest infants.
- Angiography
- Balloon
- Catheter
- Pediatrics
- Technique
Statistics from Altmetric.com
Introduction
In addition to the intrinsic complexity of cerebrovascular lesions that occur in children, safe performance of pediatric neurointerventional endovascular procedures poses unique technical challenges. Critical among these is careful consideration of the small size and fragility of both access vessels and target vessels. We describe here our extremely low incidence of femoral morbidity using rigid size parameters for arterial access, as well as two methods for safely gaining controlled distal access within the intracranial vasculature of young infants, making a new set of complex endovascular procedures feasible in this cohort.
Access size limits and morbidity
Constraints on guide catheter size as a function of patient age are critical if major morbidity from injury to the iliac–femoral arterial tree is to be avoided. Older reports, predominantly focused on infants who had undergone cardiac catheterization, describe iatrogenic femoral artery injury resulting in leg length discrepancy in 8–14% of cases, with arterial thrombosis seen in over one-third of cases1–8; complete limb loss, although rare, was also seen. Procedural details, such as the specific size of the catheters used, were often lacking in these early reports.
A study by Franken et al, systematically examining the incidence of femoral artery spasm following catheterization in a consecutive cohort of children, demonstrated a clear link between the size of the catheter used relative to the femoral artery diameter, and the likelihood of spasm.9 Burrows et al10 reviewed the incidence of iliofemoral complications in a cohort of children who had undergone catheterization for aortic or aortic valve repairs, without particular guidelines for sizing femoral catheters. They found acute iliofemoral complications in over 45% of cases, including thrombosis (28%) and arterial disruption (13%), and a correlation between the likelihood of acute complications and low patient weight. In a follow-up study,11 Burrows et al, using MRI, found that 58% of children who had undergone transfemoral balloon dilatation angioplasty had obstructive femoral artery lesions, suggesting that the catheterization had progressive deleterious effects beyond those manifest in acute postprocedural symptomatology. Critically, they found that catheter manipulation (in this case, the number of exchanges in the femoral artery) was tightly linked with the likelihood of late arterial stenosis.
In order to minimize the likelihood of arterial injury, we use table 1 in Franken et al9 as a general guide for sizing access catheters and, in every case, we use an arterial sheath that remains in place throughout the case, minimizing manipulation of the femoral access site. While numerous series have demonstrated the feasibility of glue embolization or combined coil and glue embolization through small access systems in children,12–14 complex endovascular procedures often require guide catheters with a sufficient inner diameter to accommodate devices with a larger cross section or multiple devices, and the above size restrictions ostensibly preclude a variety of treatment options. We describe here, through case illustrations, two strategies for broadening the range of usable techniques (such as balloon assisted embolization and the use of a triaxial system) in young children, via 4 F and 5 F femoral access.
Methods
With institutional review board approval, we reviewed our neurointerventional database of 268 consecutive transfemoral catheterizations in children <10 years of age, from December 2007 to June 2012 (mean age 4.7±3.2 years, median age 3.9 years); 149 of these were <5 years of age, 96 were <3 years, and 43 were <1 year. Once anesthetized, all patients underwent baseline assessment of the femoral, popliteal, and dorsalis pedis pulses in both lower extremities. Adhering to the above size constraints, we used a 4 F sheath in infants weighing <10 kg and a 5 F sheath in infants and children weighing <15 kg. All patients underwent single wall puncture of the common femoral artery using a Cook Micropuncture Introducer Set (Cook Medical, Bloomington, Indiana, USA), with ultrasound guidance for patients <2 years of age and in older patients if two attempts without image guidance were unsuccessful.
Other than the setting of acute intracranial hemorrhage, systemic heparinization was administered to all patients in the cohort, with a target activated clotting time of 200–350 s. No arterial closure devices were used in this cohort. Hemostasis was achieved with manual pressure for a minimum of 20 min and, once established, was supplemented with a Safeguard pressure dressing (Maquet, Datascope Corp, Fairfield, New Jersey, USA) inflated for 2 h (12 cm dressing for children <5 years, 24 cm dressing for older children).
Children were observed in the interventional radiology post-anesthesia care unit for at least 4 h. Lower extremity pulses, color, and temperature were checked and documented by the nursing staff every 15 min for the first hour, then every half hour for the next hour, and then every hour. Any deleterious change in lower extremity examination compared with baseline was noted and tracked, with ultrasound investigation if needed. Our protocol calls for a follow-up clinical visit 2 weeks after neurointerventional procedures, to ensure a return to neurological baseline and to review imaging findings; this follow-up visit is also an opportunity to pursue any active femoral arterial issues. For all patients whose care is centered at our institution, any clinical deficits that might be relatable to prior procedures (such as leg length discrepancy) are flagged and communicated to the involved physicians. For patients whose care is outside our center, neurodevelopmental follow-up is pursued by our nurse practitioner in consultation with the patient's physicians.
Results
Overall femoral morbidity
In our cohort of 268 pediatric procedures, there was one case of transient femoral arterial morbidity related to a stretched and sheared retained catheter in an infant after Onyx embolization of a large neck mass, as we have previously reported,15 and one case of transient femoral access site morbidity: a 3-month-old with a large choroid plexus papilloma who had undergone preoperative embolization via a right femoral approach using a 4 F system (with unremarkable post-embolization examinations) and had a left femoral arterial line placed in the operating room the next day was noted to have reduced pulses and cooler lower extremities bilaterally postoperatively. Ultrasound demonstrated moderate stenosis of the right superficial femoral artery, with diminished peak systolic velocities and compensatory dilatation of the right profunda femoris; the left side was unremarkable. The patient was followed at 2 week intervals with serial ultrasound, with a normal study at 6 weeks postoperatively. There was no leg length discrepancy or any other lower extremity abnormality on examination with over 4 years of follow-up in this patient.
No other cases of transient femoral access site morbidity were present, and there were no cases of permanent femoral morbidity in the cohort.
Case illustration: 4 F access
The patient, born at 37 weeks, was found to have a marked thrill over the scalp and precordium at birth. Echocardiography demonstrated a hyperdynamic left ventricle without heart failure. MR demonstrated a pial arteriovenous fistula supplied by massively enlarged posterior inferior cerebellar artery and superior cerebellar artery (SCA) feeders, with no evidence of brain parenchymal compromise. Genetic testing was significant for RASA1 positivity.16 Given the overall medical stability and absence of hydrocephalus, treatment was deferred until age 5 months (at which point the child weighed 7 kg). Under general anesthesia and ultrasound guidance, a 4 F sheath was introduced into the common femoral artery and the patient heparinized. A diagnostic angiogram, carried out via a 4 F vertebral catheter, confirmed the MR findings of a direct pial arteriovenous fistula, with extremely high flow fistulae draining into an ectatic venous collector and then into the torcular (figure 1). The fast flow was such that direct deployment of any embolic agent, whether solid or liquid, posed a not insignificant risk of rapid transit through the fistula into the systemic venous circulation. The patient had a patent foramen ovale, elevating the potential hazard. The decision was made to embolize the largest feeders using balloon flow arrest.
After navigating an exchange length guidewire into the right external carotid artery, the guide catheter and short groin sheath were exchanged for a 4 F Shuttle Flexor sheath (Cook Medical), with an inner diameter of 0.059 inches and a length of 90 cm. The sheath was advanced from the right common to the right internal carotid artery (ICA). A 6×9 mm Ascent Occlusion Balloon Catheter (Micrus Endovascular, San Jose, California, USA) was advanced over a Transend 0.014 inch Floppy microguidewire (Boston Scientific Neurovascular, Fremont, California, USA) into the distal left SCA via the right posterior communicating artery (figure 2A–D). Prior to balloon inflation, any coil introduced into the distal SCA, regardless of size or shape, unwound into the fistula, but with balloon inflation and flow arrest, the coil loops unfolded in a controlled fashion entirely within the left SCA. Hence, with the balloon inflated, the left SCA was embolized with two Deltapaq detachable coils (Micrus Endovascular). However, on deflation of the balloon, the coils migrated through the fistula into the torcular, where they remained. The patient remained stable, and angiography revealed no change in flow pattern. The Ascent balloon was then reinflated and in rapid succession, the left SCA was embolized with nine Axium detachable coils and one Deltapaq coil. The balloon was deflated and the coil mass stably and densely packed the left SCA with a significant reduction in flow to the fistula. The balloon microcatheter was removed and a standard microcatheter was advanced into the right SCA where direct coiling was feasible.
Given the complexity of the lesion and the age of the patient, treatment was staged. At 7 months, the infant was brought back and the remaining dominant inflow to the lesion, via the left posterior inferior cerebellar artery, was embolized directly, using detachable coils supplemented with Onyx (ev3, Irvine, California, USA). After several rounds of staged embolization, there was a dramatic overall reduction in flow to the lesion, with minimal residual arteriovenous shunting (compare figure 2E–H with figure 1A–D). The infant has remained neurologically intact throughout.
Case illustration: 5 F triaxial access
The patient was a 5 year old who presented at birth with a high flow vein of Galen malformation (VOGM) causing heart failure. Several rounds of coil embolization were performed within the first few weeks of life at an outside institution, complicated by a large intracranial hemorrhage. While the heart failure was stabilized, the patient was severely compromised neurologically, with quadriparesis and limited communication, although beloved and assiduously cared for by the family. There was significant residual flow through the VOGM and concomitant worsening hydrocephalus. Given the meaningful relationship between child and family, the decision was made to proceed with embolization in an effort to stabilize or reverse the hydrocephalus. As a result of overall failure to thrive and bedbound condition, the patient weighed less than 15 kg.
A 5 F sheath was introduced into the left common femoral artery (pulse on the right was diminished, likely from prior procedures), and a 5 F Envoy was used to perform diagnostic cerebral angiography, demonstrating a choroidal VOGM (frontal view seen in figure 3A, lateral view seen in figure 3B). The proximal cervical right ICA was found to harbor a stiff 180° horizontal curve, which did not straighten with passage of a glidewire (figure 3C). Keeping the 5 F Envoy guide catheter (Codman Neurovascular, Raynham, Massachusetts, USA) below this curve and advancing the microcatheter and microwire from this position would result in severe loss of distal control and navigability. Thus a Revive 044 Intermediate Catheter (Codman Neurovascular) was advanced through the Envoy guide catheter past the cervical kink and to the petrous ICA (figure 3D, E). An Echelon 14 microcatheter (ev3) was then introduced through the Revive and advanced over a microwire into an enlarged posterior communicating artery feeder via the posterior communicating artery, with excellent navigability (figure 3F, lateral view). The posterior communicating artery feeding pedicle was embolized with Onyx, with resultant significant diminution in flow to the lesion.
One week later, the child was brought back for further embolization. A 5 F Envoy catheter was employed to select the left vertebral artery and the Revive 044 Intermediate Catheter was advanced through the guide catheter to the V4 segment. Echelon 10 microcatheters were used to select a P1 segment feeder and a P2 segment feeder which were embolized with Onyx (figure 4A). Control angiography was carried out via the Revive Intermediate Catheter, which was then removed, and the Envoy guide catheter was selectively navigated back to the right ICA. As per the prior intervention, the Revive was used to navigate through the severe right ICA cervical kink and advanced to the distal ICA. A Marathon microcatheter (ev3) was introduced through the Revive into the distal right anterior cerebral artery pericallosal artery (figure 4B), and Onyx was used for embolization. The child tolerated both interventions without complication, remaining at his neurological baseline.
Discussion
Cerebrovascular disease in infants presents a unique challenge. Small vessel caliber and limited total blood volume make open surgical techniques extremely difficult but, conversely, minimizing iliofemoral complications in transarterial procedures requires fastidious attention to technique. Systemic heparinization, unless clearly contraindicated, is critical in lowering risk in this cohort, although specific weight based dosing strategies have not been elucidated17; it is our practice to titrate dose based on measured activated clotting time in each case. Alternative means of transarterial access are attractive in some situations as a strategy for avoiding femoral arterial complications: Berenstein et al described embolization of VOGMs in neonates by using the trans-umbilical artery route.18 However, while extremely useful in selected cases, umbilical artery access is available only in neonates (and only in cases where the intensivists have maintained patent access), and is certainly not itself free of potentially severe complications.19–21
Prior published series describing transarterial embolizations in infants are in keeping with the known predominance of VOGMs in this cohort, with pial and dural fistulae occurring less frequently,13 ,14 ,16 ,22 with excellent clinical outcome the rule, in appropriately selected patients. Access in these series was achieved via four French systems, and nearly all infants were treated with n-butyl cyanoacrylate glue embolization. A notable exception is a recent report describing embolization of 25 pediatric intracranial non-galenic arteriovenous fistulas using predominantly coils or coils and n-butyl cyanoacrylate.23 Detailed parameters regarding the size of femoral access systems are not provided in that series but the authors do describe one common femoral artery transection and one lower extremity embolus. As pointed out by these investigators, coil embolization offers the advantage over liquid embolics of precise deposition, but very high flow lesions may be best treated using flow arrest in a feeding artery pedicle, lowering the risk of systemic venous embolization. Balloon assisted Onyx embolization24 and balloon assisted coil embolization25 via 5 F systems have been described using larger systems in older patients. We describe here the feasibility of balloon assisted embolization of an arteriovenous fistula via a 4 F access system, the Shuttle flexor vascular sheath. Although not marketed for use in the carotid or vertebral arteries, this sheath's 0.059 inch inner diameter opens a range of endovascular interventions in infants that had not been previously feasible.
The introduction of two separate devices, a balloon microcatheter and a coiling microcatheter, would not have been feasible in an infant without accessing both groins and thus increasing the risk of iliofemoral morbidity. The Ascent Occlusion Balloon Catheter, a dual lumen device that offers the ability to directly embolize through the balloon catheter itself, thus proved invaluable.
The use of various strategies for shortening the distance traversed by the microcatheter to its target, particularly in cases of proximal vessel tortuosity,26–29 has become commonplace in neuroendovascular procedures. The soft Neuron System guide catheter (Penumbra, Alameda, California, USA) was originally introduced with this goal in mind. The Distal Access Catheter (DAC, Concentric Medical, Mountain View, California, USA) was developed with various diameters as an intermediate catheter introduced via the guide catheter to shorten the distance traversed by the microcatheter to its target.27 The Cleveland clinic experience with the DAC documented no complications related to its use in 103 cases, including treatment of ischemic stroke, aneurysms, arteriovenous shunts, and intracranial stent deployments.28 The Penumbra reperfusion catheters (Penumbra) were introduced soon thereafter.29
However, for procedures in young children weighing <15 kg, we limit ourselves to 5 F access systems, typically the 5 F Envoy guide catheter (Codman), with a 0.056 inch inner diameter. The DAC 044 and the Penumbra 041 have inner diameters sufficient to accommodate a wide range of microcatheters, but both of these intermediate catheters require at least a 6 F guide catheter, thus precluding their use in young children. The recently introduced Revive 044 catheter, with an outer diameter of 0.054 inches, although not marketed for use in a 5 F guide catheter system, can be effectively used in this manner, and we present here its initial application in children. Note that there is little circumferential gap between the Revive 044 and the 5 F Envoy, and care must be taken to advance and withdraw the intermediate catheter smoothly and slowly in order to prevent bubble formation.
Triaxial systems have additional advantages in the embolization of cerebral arteriovenous shunts.26 ,28 They allow for easier removal of the microcatheter after embolization without loss of distal access for subsequent embolization of additional feeding pedicles, with no need to traverse previously crossed proximal vessel tortuosity repeatedly, as illustrated here. In the context of embolization with Onyx, where the microcatheter tip must be withdrawn from a proximal plug, intermediate catheters ensure that less torsion is applied to the length of the arterial tree. Additionally, given their larger inner diameters, intermediate catheters may also prove useful in themselves as diagnostic catheters in lieu of small microcatheter injections (figure 4D, F).
Conclusion
In a cohort of 268 procedures done in children <10 years of age, we report a single case of transient femoral access morbidity and no permanent morbidity. We present here methodologies for achieving safe distal intracranial access for neuroendovascular procedures in infants, while respecting size limitations on femoral access. Neither the 4 F Shuttle sheath nor the Revive 044 Intermediate Catheter was developed for use in children, and in fact virtually no neuroendovascular devices are developed or marketed specifically for pediatric indications. Thus borrowing creatively from the adult neuroendovascular armamentarium is a necessity in treating children. No doubt new devices with potential use for pediatric procedures will be developed and released in short order, and it behooves the alert practitioner to keep an open mind towards adopting such tools on an ongoing basis.
References
Footnotes
-
Contributors BAG wrote the first draft of this paper and collected the figures. DBO conceived of the project and edited and partially rewrote the manuscript and figures.
-
Competing interests None.
-
Ethics approval The study was approved by the Boston Children's Hospital institutional review board.
-
Provenance and peer review Not commissioned; externally peer reviewed.