Robotic and Navigational Technologies in Endovascular Surgery

VASCULAR DISEASE MANAGEMENT 2010;7(1):E15-E19

Introduction

Endovascular treatment of arterial disease has been a revolution for clinicians and patients alike, allowing treatment of aortic and peripheral pathologies percutaneously or through small incisions, reducing the morbidity and mortality associated with open surgical treatment of vascular disease. Moreover, it has allowed the treatment of patients whose respiratory and cardiovascular comorbidities would have rendered them too high risk for open surgery. Endovascular approaches have become more widely adopted as the decrease in morbidity and mortality have been recognized in clinical trials.^1,2 Endovascular treatment, however, is by no means a panacea for the treatment of all arterial disease. The anatomical suitability of patients for endovascular aortic aneurysm repair must be closely scrutinized. Features such as the caliber of access vessels, tortuosity and angulation of the aorta, length of landing zone and involvement of visceral vessels, must all be taken into account, and procedures with difficult anatomical configurations can be technically challenging. Conventional catheters have a limited range of shape, flexibility and manoeuvrability, and the mainstay of image guidance in most endovascular suites remains two-dimensional (2-D) fluoroscopy, with its attendant radiation exposure to patient and staff and the need for nephrotoxic contrast. For these reasons, we have been working to find solutions to the current limitations of endovascular surgical tools, examining the use of robotically controlled endovascular catheters and alongside this, we are investigating novel imaging platforms for improved endovascular navigation. Our aim is to couple intuitive three-dimensional (3-D) imaging, which ideally would involve neither fluoroscopy nor nephrotoxic contrast agents, utilizing devices that can be steered remotely and accurately within the arterial tree.

Robotic Technology

The use of robotics has previously been adopted in cardiac and urological surgery in which master-slave systems such as the Da Vinci® robot (Intuitive Surgical, Sunnyvale, California) allow for greater precision of movement through an intuitive hand-operated console which is able to translate fine movements of the fingers into exact actions performed on the target tissue. This system has been in use at our institution since 2001 and has been used in radical prostatectomies³ and over 130 coronary artery bypass procedures but, as yet, does not offer a suitable platform for endovascular procedures. Within the vascular tree, a potential role for robotic technology is to overcome some of the current limitations of conventional endovascular catheters: conventional catheters have a relatively small repertoire of shapes and sizes and, more importantly, lack active maneuverability of the tip and stability during wire exchanges and passage of other endovascular tools. This necessitates frequent catheter changes, which is time-consuming, threatens the position of guidewires, and may result in distal embolization or vessel trauma. This is particularly pertinent during instrumentation in the aortic arch, where cerebral embolization may occur, resulting in stroke. The limitations of traditional catheters become particularly apparent when attempting to cannulate visceral vessels during placement of fenestrated stent grafts for the exclusion of perivisceral aneurysms. Such stent grafts are currently manufactured on a bespoke basis and have had good results reported in the short and medium term.^4–6 However, the implantation of these devices is technically demanding. The process of cannulating branch vessels through the fenestrations is extremely technically challenging and is often complicated by other anatomical issues. These include: tortuous iliac arterial systems which can interfere with catheter manipulation; the presence of thrombus and calcification within the vessels which, combined with multiple attempts at vessel cannulation, may lead to distal embolization; aberrant renal vessels; and finally, tortuosity and angulation of the aorta above or at the level of the fenestrated segment which can lead to graft rotation and significant misalignment of fenestrations with vessel ostia preventing straightforward target-vessel cannulation. There are commercially available catheters available with manually “shapeable” tips and steerable sheaths or guide catheters. These devices have been developed in an attempt to overcome the problems outlined above. However, they are heavily dependent upon the physician’s skill, as he or she must either manually shape the catheter tip prior to cannulation, or use the catheter’s pull-wires to achieve the appropriate tip curvature. The physician must then apply torque and translational action to maneuver the tip into the target vessel. Emerging technologies utilizing magnetic fields offer the ability to remotely control the catheter tip. These include the Niobe® (Stereotaxis, Inc., St. Louis, Missouri). In this system, which is currently used for the ablation of arrhythmogenic cardiac pathways, two permanent magnets, whose location can be controlled by a central computer, generate a magnetic field of around 0.08 T around the patient’s chest. The distal tip of the mapping or ablation catheter contains three magnets that automatically align with the direction of the external magnetic field, thereby steering the catheter.⁷ Such a magnetic navigation system has been reported in the treatment of human peripheral arterial disease.⁸ The disadvantages of this system are the cumbersome nature of the magnets and the time delay between instruction and movement of the catheter tip. The magnetic field produced also requires a specially built room. We have installed in our aortic and peripheral endovascular suite a Hansen Sensei robotic system (Hansen Medical, Mountain View, California) (Figure 1), the first installed for this purpose worldwide. This “master-slave” system allows the physician to remotely control and position steerable catheters within the vascular tree. The robotic catheter (Artisan) is comprised of a flexible inner guide (11 Fr outer diameter, 8.5 Fr inner diameter) within a steerable, but stable, outer guide (outer 14 Fr, inner 11 Fr). The physician is seated at a remote workstation away from the patient and therefore out of the radiation field. The control workstation uses electromechanics to translate the movements of the physicians hand on the 3-D mouse into accurate movements at the tip of a guide catheter and sheath (Artisan catheter), enabling 3-D control of the catheter tip with seven degrees of freedom. The catheter is navigated using a console displaying pre-operative computed tomographic (CT) images, intra-operative fluoroscopy and a Cartesian 3-D representation of the location of the catheter tip (a virtual catheter). As seen in laparoscopic surgery, a major limitation of remote catheter control systems is the lack of tactile feedback, which makes it difficult for the operator to gauge the amount of force that is being applied to the target tissues. A possible simple solution to this is to offer a visual surrogate. One example of this concept is the Intellisense™ Fine Force Technology (developed by Hansen), which graphically displays the amount of force being applied to the catheter tip on the physician’s console (Figure 2). This safety feature may alert the physician to the danger of continued force and decrease the likelihood of inadvertent damage to the vessel wall, or even perforation.

Imperial Experience with Robotic Endovascular Systems

We have utilized phantom, pulsatile flow models to gain familiarity with the Hansen Sensei system, to train our experienced endovascular operators in its use, and subsequently, to study its effect on performance during simple and complex vessel cannulation tasks (Figure 3). Our users have adapted very quickly to this new technology. They have reported that the maintenance of stable positioning conferred by the robot and the intuitive translation of hand movements into catheter-tip movements are the greatest attributes of such a catheter system. Objectively, we have found that there was no clear benefit seen over conventional methods in the performance of simple tasks, but during complex cannulation procedures, this translated into shorter times for vessel cannulation, fewer tool movements performed to achieve each task, increased accuracy in cannulation of vessels, and improvement in performance scores.^9,10 Following in-vitro and animal model training, in August 2008, we were able to undertake the first human robotically assisted endovascular aneurysm repair¹¹ using the robotic catheter to cannulate the contralateral limb of an infrarenal stent graft. This was a “world first” for endovascular robotics and, as such, clearly represents a breakthrough in vascular intervention. However, metrics which we have collected on vessel cannulation times and number of movements seen using this system in the in-vitro setting, demonstrate that for operators who are already experienced in conventional endovascular work, its benefit lies not in simple tasks such as contralateral endovascular aortic aneurysm repair (EVAR) limb cannulation, but in more complex aortic procedures such as fenestrated EVAR and arch-vessel cannulation.^12,13 We postulate that in human procedures, these metrics should be reflected in decreased total procedure time, decreased fluoroscopy time, and reduction in the volume of contrast agent required, with improved safety for patients and clinicians. Enhanced catheter stability for the introduction of stents, guidewires and other endovascular tools may reduce the need for multiple cannulation attempts and, consequently, the risk of damage to the target vessel. Fenestrated EVAR, which is currently a complex and time-consuming procedure, may be simplified by this robotic technology, allowing us to treat a greater number of patients via a totally endovascular approach. Other potential applications of such a platform include use in carotid intervention and the treatment of complex peripheral arterial disease: the enhanced stability and maneuverability of a robotic catheter could improve the outcome of endovascular treatment of complex lower-limb occlusive disease and could reduce the risk of cerebral embolization during cannulation and stenting of the internal carotid artery. However, miniaturization of the technology, allowing a reduction in size of the robotic catheter from the current 14 Fr, would first be necessary. In our experience, acclimatization to the use of this robotic system is easily achieved by both experienced and novice operators. Despite a short training period on the robotic system, during in-vitro trials, experienced operators reach proficiency quickly, whereas the less experienced operators can improve their performance to a high standard within the same time period^.9,10 In addition, the robotic system may assist in achieving in-situ fenestration of stent grafts, which would obviate the need for expensive, bespoke fenestrated stents, which are costly and can take many weeks to manufacture. The principle of in-situ fenestration is the use of a generic main body stent through which punctures are created from the aortic side and visceral stents placed in the target vessels. To date, we have achieved antegrade in-situ fenestration in porcine models using the robotically steered catheter system and 3-D rotational angiography. Accurate and stable positioning of the catheter tip inside the stent material and adjacent to the vessel ostia are crucial in this procedure in order to allow puncture of the graft and passage of a side-branch stent. This proposed technique of robot-assisted in-situ fenestration may offer the option of total endovascular aneurysm treatment to many patients who are either anatomically unsuitable for custom endovascular fenestrated stent grafting, or who require the procedure on an urgent basis.

Advances in Imaging and Navigation

While stent-graft technologies have progressed, the mainstay of image guidance in most endovascular treatment suites remains 2-D fluoroscopy requiring the administration of contrast which is potentially nephrotoxic. Many centers now have access to on-table CT, such as the Siemens DynaCT, however, this does not offer real-time intraoperative localization of endovascular tools and without this, remotely controlled endovascular procedures are unable to achieve their full potential. There are several novel, real-time, 3-D imaging solutions available, developed mainly for neurosurgery and cardiology. These systems have demonstrated the ability to enhance the accuracy of procedures such as the excision of intracerebral lesions and the ablation of accessory cardiac pathways. Several systems use electromagnetic technology to map the surrounding anatomy and track the active tool tip to its desired location. One particular system is the Stealth® Axiem™ which utilizes an electromagnetic field generated around the patient, combined with a tracking electrode and instruments with a single magnetic coil at their tip to calculate the position of the instrument tip. The single-coil technology is lightweight and unobtrusive, allowing very flexible, slim tools to be tracked. This electromagnetic tracking data is then registered with reconstructed, preoperative CT or magnetic resonance imaging (MRI) images, which are mapped to the patient, allowing movements of the catheter tip to be traced in real time on these 3-D reconstructed images. The 3-D image for navigation and dynamic cross-sectional views of the anatomy in the three Cartesian axes are displayed on the monitor. To date, this system has been widely used in orthopedic surgery and neurosurgery with promising results. In the neurosurgical setting, the system has also been used with ultrasound technology (SonoNav™ and SonoSite™, Medtronic Surgical Navigation Technologies, Minneapolis, Minnesota) to assist the clinician in locating brain tumours and to compensate for the brain shift, which occurs during craniotomy. At our institution, we have found the system to be a viable navigation method in in-vitro aortic phantom models and are working to expand its application in this field (Figure 4). There are other operating systems in use in the field of interventional cardiology, which also utilize electromagnetic technology and combine it with intrinsic cardiac electrical activity. Endocardiac mapping can be achieved to identify and perform ablation of arrhythmogenic pathways. The Carto™ System (Biosense Webster, Inc., Diamond Bar, California) performs navigation and remote mapping of the cardiac chambers. It uses an external, ultra-low magnetic field-emitter pad, which is placed on the patient and its signal is detected by a magnetic sensor located on the catheter just proximal to the mapping electrodes which record intrinsic activity from the cardiac muscle. A processing unit receives this information and translates it into a 3-D anatomical and functional cardiac image which can be used to guide ablation in real time. This basic cardiac image can then be married with pre-operative CT imaging to enhance the image quality for navigation. This can also be combined with intracardiac echocardiography. A similar system used in cardiology is the St. Jude EnSite® NavX™ system (St. Jude Medical, Inc., St. Paul, Minnesota,), which uses a slightly different approach to cardiac mapping and ablation: the use of electrical currents. Electrical signals transmitted between three pairs of surface electrodes are sensed by standard electrophysiology catheters and fed back to the central processing unit, which utilizes the information to construct a 3-D cardiac model. The system has the advantage of being able to display multiple catheters at any one time (up to 12) and has additional software that allows it to be combined with CT data. This technology has shown promising clinical results for cardiac pathway ablation. Another alternative approach to navigation is the combination of preprocedural cardiac MRI, coupled with intra-operative MR fluoroscopy. This is reported to achieve very detailed 3-D anatomy without the need for any specific mapping technology such as those already described. In the vascular setting, MR fluoroscopy has been reported in the placement of an aortic stent graft in an in-vitro, non-pulsatile aortic phantom.^14–16

Combined Navigation and Remote Control

The ultimate goal in endovascular surgery is certainly to marry navigation with remote-control catheter technology, something which has also already been achieved with success in the cardiac field in which the Carto™ mapping system (Biosense Webster) has been used in conjunction with the Niobe® remote magnetic navigation system described previously. This offers some exciting features such as automatic navigation of the catheter to a target point selected by the physician on the mapping system.⁷ These cardiac navigation systems can also be used with the Hansen robotic technology. The benefits of combining these approaches, as perceived by the clinicians currently using them, are: precision of catheter positioning beyond that possible with conventional means and a consequent decrease in procedure time, fluoroscopy time, and overall safety and efficacy of the procedure.

Other Developments

Mathematical approaches have also been made in the area of navigation: researchers at Siemens, for example, have established a statistical framework which, using traditional 2-D fluoroscopy, can compute a reliable estimate of the location of a guidewire tip within a 3-D vessel model.¹⁷ There is still much work to be done to achieve the ideal imaging platform for the navigation of remotely controlled endovascular procedures. An ideal solution would offer 3-D, real-time imaging and would remain accurate despite respiratory motion and pulsatile aortic movement. It would offer a high-quality endoluminal view, as well as visualization of surrounding tissues and would delineate aortic pathology such as thrombus or calcification. It would eliminate the need for radiation or potentially harmful contrast administration and should be versatile, simple to set up and importantly, capable of accurately registering a variety of endovascular tools. There are clearly aspects of the imaging technology described above which go some way towards meeting these needs and, on that basis, they warrant further investigation in an endovascular robotic setting refined for use specifically in the vascular tree.

Conclusions

Technological advances have allowed the endovascular treatment of abdominal aortic aneurysms to become widespread, offering a safer alternative to open surgery and allowing intervention for patients who would otherwise have been deemed inoperable. Many patients, however, are not suitable for endovascular intervention. We anticipate that robotic technology, assisted by navigation techniques, will simplify endovascular tasks and increase the applicability of minimally invasive technology to a greater patient population. At present, robotic technology for vascular intervention is at a point where it can realistically be utilized to treat selected patients under controlled circumstances. It demonstrates the potential for improved accuracy and stability of endovascular tool placement and, we believe, with further development, it will also confer safety for patients and staff. In conjunction with navigation technology, robotic technology holds a great deal of promise for the treatment of complex vascular disease.

From the *Department of Biosurgery and Surgical Technology, Imperial College of London, the §Department of Vascular Surgery, Imperial College Healthcare NHS Trust, and the £Department of Interventional Radiology, Imperial College Healthcare NHS Trust, London, United Kingdom.

Disclosures: Dr. Bicknell discloses that Medtronic has provided cliical lectureship funding to his institution; Dr. Cheshire is on the executive board of Veryan Medical. The other authors report no potential conflicts of interest regarding the content herein.

Address for correspondence: C. Bicknell, MD, Senior Lecturer in Surgical Technology and Consultant Vascular Surgeon, Imperial College, Vascular Secretaries Office, Waller Cardiac Building, St. Mary’s Hospital, Praed Street, London W2 1NY, United Kingdom.