ADVERTISEMENT
In-vitro Model of Translumbar Embolization of Endoleaks with NBCA: Risk of "Gluing" Different Access Devices
Abstract
Endoleaks are well-reported complications of endovascular aneurysmal repair, with the type 2 varieties being the most common. Repair of this complication is performed via transarterial or percutaneous translumbar approaches, and use of N-butyl cyanoacrylate (NBCA) as the embolic agent has shown promise. There is potential, however, for adherence of catheters and devices to this liquid embolic agent. Objective. We wanted to test the utility of a flowing model to simulate translumbar access to treat endoleaks. We sought to assess the risk of adherence of different devices used for translumbar injections of NBCA at different concentrations. Methods. An in-vitro flowing model simulating an aneurysmal sac with endoleak was created. Swine plasma was circulated into a simulated sac with one entry and two exit points. Various access devices were used to access and embolize the aneurysm sac with NBCA, and the resistance to remove the devices was studied. Results. Of the devices tested, no resistance was felt when withdrawing metallic needles and plastic sheaths. Mild to no resistance was noted for catheters, and only the sheath with a curved catheter became adhered to the aneurysm sac (balloon). Conclusion. An in-vitro test of flowing translumbar access was successfully created using simple and inexpensive devices. In this model, the risk of “gluing” the devices in place was low. Metallic needles and straight sheaths did not have any resistance, and reflux of a concentrated mixture of embolic agent resulted in adherence of the curved catheter and sheath.
Introduction
Introduced by Parodi et al in 1989, endovascular aneurysm repair (EVAR) using a stent-graft has gained widespread acceptance and dissemination since the US Food and Drug Administration (FDA) approval of commercial devices in 1999. A well-reported complication of stent-graft repair is persistence of blood flow outside the graft lumen, defined as an endoleak. Endoleak rates established from single institutional studies and follow up from original clinical trial patients vary from 15–50%.3 These rates remain controversial. A classification scheme has been established to describe types of endoleaks due to various etiologies. There is general consensus that endoleaks of type 1 (attachment site) and type 3 (graft defects or separations) need to be treated urgently.12 There is no such agreement on the treatment of type 2 endoleaks, whose incidence varies greatly in literature and has been reported to be as high as 30%.5 In general, treatment is advocated for endoleaks with enlarging aneurysm sacs and persistent type 2 endoleaks regardless of aneurysm sac diameter (i.e., endoleak persistence for greater than 12 months).5 Treatment has comprised of transarterial and translumbar approaches, traditionally using coil embolization. More recently, liquid embolic agents, such as N-butyl cyanoacrylate (NBCA, Cordis Neurovascular, Miami Lakes, Florida) have been used alone or in combination with coils for the management of stent-graft endoleaks via translumbar approach. Cyanoacrylates were initially described as liquid embolic agents for the treatment of cerebral arteriovenous malformations and arteriovenous fistulas. They have been in clinical endovascular use as tissue adhesives since the 1970’s. Original clinical experience was gained with a vinyl monomer of alkyl 2-cyanoacrylates, isobutyl 2-cyanoacrylate (IBCA).10 This agent became unavailable in the 1980’s, and was replaced by a similar compound, N-butyl cyanoacrylate (NBCA).9 NBCA was approved by the FDA for use specifically in the treatment of intracranial aneurysms by 2000.9 Since then, other uses of this agent have been found in the treatment of intra-abdominal and peripheral arteriovenous malformations/fistulas, pseudoaneurysms, biliary complications, acute arterial hemorrhage (used in some cases for hemorrhage refractory to coil or particle embolization), and management of endoleaks.9,17 Some of the reported complications of NBCA embolization include abscess formation, non-target embolization, and rare eventual recanalization of embolized vessel. Hemorrhagic complications are reported during AVM embolization as being due to microcatheter/microwire manipulations, or rupture of small vessels during contrast and/or embolic agent injection. Another classically reported complication includes entrapment of the delivery catheter within the embolized artery. As clinical applications of NBCA embolization for treatment of various pathologies have increased, access devices other than endovascular microcatheters are being more commonly used in clinical practice for targeted delivery of this embolic agent. Prior reports have studied the risk of adhesion of NBCA glue to various hydrophilic and non-hydrophilic microcatheters used for intracranial endovascular embolization.10 These can be serious complications where the microcatheter is “glued” in place to the intra-cerebral vessel of the patient warranting surgery and/or anticoagulation to prevent thromboembolic sequelae.8 However, there have been no studies studying the risk of adherence of glue to other access devices, including sheaths, needles, and catheters for use in non-endovascular procedures, such as translumbar embolization of endoleaks. In theory, adhesion of the devices to the patient’s tissues or organs is possible if the concentration of the NBCA is too high and/or the exposure of the catheters is prolonged. We devised an in-vitro model to assess the risk of “gluing” various access devices such as needles, sheaths, and routine catheters, which can potentially be used specifically for translumbar injections of NBCA at different concentrations. We also tested the utility of a flowing model using swine plasma to simulate translumbar access to treat endoleaks.
Materials and Methods
The following access devices and catheters were used in this study: Accustik sheath (Boston Scientific, Natick, Massachusetts), 6 Fr sheath (Arrow Intl., Reading, Pennsylvania) with Omniflush catheter (AngioDynamics, Queensbury, New York), 4 Fr micropuncture sheath (Angiotech, Vancouver, British Columbia), 4 Fr Skater sheath (Angiotech), 5 Fr sheath (Terumo Interventional Systems, Somerset, New Jersey), 5 Fr hockey-stick shaped catheter (AngioDynamics), 5 Fr SOS OMNI catheter (AngioDynamics), 5 Fr Cobra catheter (AngioDynamics), 18-gauge needle (Angiotech), 20-gauge needle (Angiotech). Trufill NBCA (Cordis Neurovascular) mixed with pre-packaged ethiodized oil (Ethiodol, Savage Laboratories, Melville, New York) (concentrations of 1:1 or 1:2) was used as the liquid embolic agent. An in-vitro apparatus simulating an aneurysm sac with endoleak was created with a balloon of a Safeguard pressure-assisted device (Datascope Corp., Montvale, New Jersey). The balloon was attached to two plastic tubings on either side, thus creating a flow model with one entry and two exit points. This simulated an endoleak sac with one feeding and two draining vessels. The device was attached to a RFA pump (RITA Medical Systems Inc., Fremont, California) with continuous circulating flow of swine plasma, thereby translating hydrostatic pressure into the balloon. No additional pressure to the balloon was added. Puncture of the balloon was performed under direct visualization and fluoroscopy to emulate percutaneous translumbar access. Various access devices and catheters listed above were individually tested. Each device was used to puncture separate Safeguard device-balloon sacs. A 5% dextrose (D5W) solution was flushed through the access device, after which the device was occluded to prevent reflux of swine plasma. Two different concentrations of NBCA glue/Ethiodol mixture (1:1 and 1:2) were injected into the balloon sac, until the two exit points were successfully embolized and sac completely filled with embolic agent. For one device (Accustick sheath), a NBCA/Ethiodol concentration of 1:3 was used. The access devices were left in place for 60 seconds and withdrawn from the embolized balloon. Resistance to remove the devices from the balloon sacs was qualitatively measured.
Results
A total of 11 fluoroscopic injections of NBCA glue were performed at different concentrations. There was successful embolization of the two exit points in the majority of the injections. Two injections filled the sac completely without occlusion of the exit points. There was leakage of the plasma noted at the attachment sites of the tubes to the balloon, due to excessive pressure of the system once the exit points were completely occluded. The flowing model was adequate to test the injection of different concentrations of the mixture under fluoroscopic guidance. There was no resistance to withdrawing the Accustik sheath, 4 Fr micropuncture sheath, 4 Fr Skater sheath, 5 Fr Terumo sheath, 18-gauge needle, and 20-gauge needle. The devices demonstrating mild resistance to withdrawing after injection of glue included the 5 Fr hockey-stick shaped catheter (both 1:2 and 1:1 glue concentrations), and 5 Fr Cobra catheters. The only device which became glued in place was the 6 Fr sheath with Omniflush catheter (glue concentration of 1:1). There was significant reflux noted of the glue inside the sheath and around the catheter, with extrusion of the glue around the entry point in the sac. The other catheters and sheaths did not demonstrate this complication. Using this model, we determined that the risk of adherence of devices to NBCA is low. Metallic needles and straight sheaths did not have any resistance. Adherence of the curved catheter and sheath (6 Fr sheath with Omniflush catheter) to the glue was due to significant reflux of very concentrated NBCA around the catheter.
Discussion
Type 2 endoleaks occur at a rate of 5–25% after EVAR,6 and can be as high as 30%, according to some literature.5 This complication has been attributed to blood flow taking a circuitous route, travelling via collateral branches from non-covered portions of aorta or iliac vessels, and transmitting arterial pressure into the aneurysm sac. Early or primary type 2 endoleak is defined as one that is diagnosed during initial imaging follow up, and late or delayed is defined as endoleak diagnosed after 30 days of follow up with a prior negative CT scan.7 Persistent endoleaks are defined as those lasting longer than 6 months.7 Spontaneous sealing can be expected in 58% of endoleaks up to 12 months. However, after 12 months of nonresolution, the endoleak is unlikely to spontaneously seal.5 Persistent endoleaks are also associated with many adverse effects, such as higher risk of aneurysm sac growth, need for conversion to open repair, and increased re-intervention rates, and are a significant predictor of aneurysm rupture.7 Therefore, some institutions employ a more aggressive approach in terms of re-intervention if a type 2 endoleak persists for more than 6 months, with mandatory treatment after 12 months.5 The treatment of type 2 endoleaks can be performed via a transarterial approach or direct translumbar aneurysm puncture, with the use of coils, glue, or a combination of both. Technical success of the translumbar technique was first reported by Baum et al (2001), who successfully treated 6 out of 7 patients (3 of whom had failed transarterial therapy). Baum et al, in 2002, reported that direct translumbar embolization is effective in elimination of type 2 leaks and should be used for aggressive management, rather than transarterial embolization. A transarterial approach accesses the culprit collateral vessel endovascularly to deploy metallic coils for embolization near its communication to the aneurysm sac. This method has limitations in that it can be difficult, time-consuming, and is not possible in all patients, due to limitations posed by anatomy. The study by Baum et al showed that 16 of 20 (80%) of type 2 endoleaks treated transarterially recurred. They postulated that the high recurrence rate is due to an endoleak being a complex vascular structure containing a nidus with multiple feeding and draining vessels, similar to a vascular malformation. Embolization of the nidus is essential for adequate treatment.1 The study demonstrated that 12 out of 13 (93%) endoleaks treated percutaneously were successful and durable (with a median follow up of 254 days).1 Percutaneous translumbar embolization obviates the need to traverse cavities or organs, thereby minimizing risk of complications. It aims to occlude the endoleak nidus itself and prevent communication between aortic branch vessels and aneurysm sac, very similar to nidus embolization in arteriovenous malformations.15 As these endoleaks are dynamic structures resembling arteriovenous malformations fed by multiple feeding and draining vessels, identification of collateral vessels may not be mandatory for procedure success.6 Our flow model simulated this very scenario of an endoleak behaving like an arteriovenous malformation. Liquid embolic agents, such as NBCA, have recently gained in popularity as the embolic agents of choice for this procedure. One of the potential risks of injection of NBCA is the adherence of various access devices to the agent in the vessel or structure being embolized. This complication is detailed in the article by Mathis et al, and can be due to factors such as polymerization time, glue viscosity, and catheter material. The authors created an in-vitro model comparing various cerebral microcatheters with and without hydrophilic coatings. Their findings indicated that a weaker glue-catheter bond forms between NBCA and hydrolene (the chemical coating utilized to transfer hydrophilic properties to microcatheters) than with the non-hydrophilic catheter materials.10 The complication of catheter adherence is also described in a technical case reported by Kelly et al, where a cerebral microcatheter adhered to a newer liquid embolic agent, Onyx (ethylene-vinyl co-polymer, ev3, Irvine, California) during dural arteriovenous fistula embolization in the right occipital artery. The procedure entails creating an occlusive cast using the liquid embolic agent around the distal tip of the microcatheter, so that a vascular plug is created. The embolic agent can then be injected to achieve distal penetration into the nidus of the arteriovenous fistula while minimizing reflux. This same principle is used for embolization using NBCA. If the microcatheter becomes adhered to the glue, it can stretch and fracture, resulting in a retained catheter, which may need to be removed surgically. If surgical removal is not possible, the patient will likely require anticoagulation or antiplatelet prophylaxis, as thromboembolic complications can occur.8
Not well documented, however, are the potential risks of adherence of various access devices used in translumbar endoleak repair procedure to NBCA. Our experiment sought to create an in-vitro model to study this scenario. Access devices used commonly during translumbar embolization were individually tested. An endoleak was simulated using a balloon of a Safeguard pressure-assisted device. The balloon was attached to two plastic tubings on either side and the tubings were then attached to a small pump, which emulated an endoleak with feeding and draining branches. A flow loop was created, through which swine plasma was circulated. This, in effect, simulated a mini-circulatory system, translating flowing plasma pressure into the balloon. The only device which became glued to NBCA within the balloon was the 6 Fr sheath with Omniflush catheter. Mild to no resistance was encountered with the other devices. The reason for adherence of 6 Fr sheath with Omniflush catheter was a significant reflux of glue inside the lumen of the sheath and around entry points. The Omniflush catheter is a curved catheter and likely also exhibited increased frictional properties during the withdrawal attempt. Adherence of this catheter was also due to injection of very concentrated NBCA (1:1). Needles did not have any risk of adherence to glue, likely due to their metallic composition. Overall, the risk of “gluing” devices in place was very low, even in this model where very concentrated NBCA mixtures were used and the devices were left in place for a prolonged time of 60 seconds. Aside from using specific access devices with the lowest risk of adherence to NBCA, manipulation of glue properties is also helpful in reducing complication rates during a translumbar embolization procedure. Polymerization times of NBCA, for example, can be adjusted by the addition of iodinated oils, such as Ethiodol, iophendylate (Pantopaque), or glacial acetic acid.16,17Polymerization times ranging from 0.2 to 5 seconds can be achieved with certain mixtures.10 Continuous injection of D5W through the guiding catheter and flooding the targeted vessel with this non-ionic solution also allows NBCA to have increased polymerization times.4 This technique has been used for distal penetration of glue into the target vessel, and can be applied for embolization of nidus and feeding/draining vessels of an endoleak. For our study, access devices were also flushed with D5W before each glue injection. Vessel tortuosity of the AVM also plays a role in adherence. There will be increased resistance to withdrawl from the glue cast in a tortuous feeding or draining branch vessel of an AVM. For translumbar embolization of endoleaks, no such vessel tortuosity is encountered, as the access device is guided directly to the endoleak nidus in a linear path. This is likely another reason for low risk of device adherence to NBCA during translumbar embolization.
NBCA is currently approved by the FDA for pre-surgical embolization of cerebral arteriovenous malformations.13,14 Technical success of NBCA utilization as the liquid embolic agent for treatment of type 2 endoleaks via translumbar approach was reported by Stavropoulos et al. In this study, 9 patients with an average age of 82 years were treated for 11 type 2 endoleaks and followed over a 9-month period. All patients had diagnosis made by a follow-up CT angiography. NBCA glue alone or in combination with coils were used, and 6 out of 9 patients (66%) had complete embolization on immediate and subsequent follow-up imaging. Persistent endoleak was demonstrated in 3 patients, who were then successfully treated with repeat translumbar NBCA injection alone.12 In a more recent study by Gorlitzer et al, 17 out of 84 patients who underwent EVAR from 2001–2007 were determined to have type 2 endoleaks. Translumbar embolization was used for treatment in 5 of these patients, using Onyx liquid embolic system in combination with cyanoacrylic-based surgical glue (Glubran), similar to NBCA. The procedure was tolerated by all patients without complications related to sac puncture. This study demonstrated the feasibility and efficacy of translumbar access to treat type 2 endoleaks using liquid embolic materials.6 By testing various access devices used in translumbar treatment and their potential for adherence to NBCA, our study helps reaffirm the low risk of treating this EVAR complication.
Conclusion
Type 2 endoleaks are associated with many adverse effects, such as higher risk of aneurysm sac growth, conversion to open repair, and increased re-intervention rates. They are a significant predictor of aneurysm rupture.6 It has been established in literature that translumbar embolization to treat endoleaks is an effective and well-tolerated procedure with a relatively low risk of complications. Its effectiveness is, in part, due to obliteration of the central nidus of the endoleak, analogous to arteriovenous malformation embolization.15 Liquid embolic agents such as NBCA and Onyx have grown in favor as embolic materials of choice compared to metallic coils, but carry a potential risk of adhering to various access devices used to inject them.
Our study demonstrated that the risk of “gluing” devices in place is very low in this experimental model for translumbar glue-embolization procedures. This study also showed that an in-vitro model can be created to adequately test the injection of different concentrations of the mixture using various access devices utilized in translumbar approaches. We believe that use of this model can be helpful in preprocedural planning of glue embolization with an access device of choice not already tested in this study. The model can also potentially be used to test other embolic agents, such as the Onyx liquid embolic system. Our model did not simulate transarterial endoleak embolization technique, thus catheter adherence to glue during this particular procedure was not tested. A study modeling this scenario should also be devised and undertaken, because the procedure is often utilized and current literature on the topic is minimal.
From University of Texas Health Science Center at San Antonio, Department of Diagnostic Radiology.
Manuscript received June 1, 2009, provisional acceptance given June 29, 2009, accepted July 7, 2009.
Address for correspondence: Jaideep Barge, MD, University of Texas Health Science Center San Antonio — Dept of Radiology, 11030 Rindle Ranch Road, San Antonio, TX 78249.
Disclosure: The authors report no financial relationships or conflicts of interest regarding the content herein.
