Novel diagnostic and imaging techniques in endovascular iliac artery procedures

ABSTRACT

Introduction

Endovascular revascularization has become the preferred treatment for most patients with iliac artery obstructions, with a high rate of clinical and technical success.

Areas covered

This review will describe novel developments in the diagnosis and treatment of iliac artery obstructions including the augmentation of preprocedural imaging with advanced flow models, image fusion techniques, and state-of-the-art device-tracking capabilities.

Expert opinion

The combination of these developments will change the endovascular field within the next 5 years, allowing targeted iliac treatment without the need for radiographic imaging or iodinated contrast media.

KEYWORDS:

• Angiography
• contrast media
• iliac artery disease
• image fusion
• pressure measurement
• radiation dose
• stenosis
• tracking

1. Introduction

In up to one-third of patients with peripheral arterial occlusive disease (PAOD), the iliac arteries are involved [1]. Since the first angiogram [2], technological improvements in endovascular procedures have been substantial. Endovascular revascularization has now become the preferred treatment for most patients with iliac artery obstructions [3] suffering severe claudication or chronic limb-threatening ischemia.

Optimal treatment planning using computed tomography angiography (CTA), magnetic resonance angiography (MRA), or a combination of the two, is of paramount importance. Despite innovations in imaging and therapeutic options, some hurdles still have to be taken to optimize treatment of iliac artery disease. In the majority of patients a hemodynamic significant stenosis or occlusion can be seen on diagnostic imaging, but in the case of equivocal iliac artery stenoses or multilevel disease, the diagnostic accuracy of current imaging modalities is not always sufficient to translate lesion diameter reduction (or stenosis) into hemodynamic parameters such as the translesional pressure gradient [4,5].

Treatment planning solely based on standard imaging could lead to both under- and overtreatment of equivocal stenoses. Overtreatment of hemodynamically nonsignificant stenosis can potentially lead to higher costs and unnecessary complications, whereas treatment of these lesions is unlikely to improve clinical symptoms [6]. In these circumstances, enhancement of standard imaging with functional assessment by calculating the pressure gradient across an equivocal iliac artery stenosis and predictive modeling [7–11] could play a major role in a targeted treatment strategy for the individual patient.

In addition to this, radiation exposure remains an issue requiring attention due to the possible negative effects on patients and endovascular specialists as well as the effects of iodinated contrast media (CM) use on patients. Imaging protocols to reduce radiation and the amount of iodinated CM have already been globally implemented, but further improvements are mandatory [12–18].

Image fusion guidance has been introduced to reduce radiation exposure and the amount of iodinated CM [19]. With image fusion guidance, 2 single-shot orthogonal exposures or periprocedurally acquired three-dimensional (3D) cone-beam computed tomography (CBCT) images are fused with preprocedural cross-sectional CTA or MRA imaging. Techniques using image-guided fusion have already proven to reduce the use of iodinated CM in endovascular aneurysm repair (EVAR) [19–22], but not in iliac procedures.

Traditionally, 3D fusion is used together with fluoroscopy, but more recently, other imaging and visualization techniques, such as fiber optic technology and electromagnetic tracking [23–26], are finding their way to the angiography suites and hybrid endovascular rooms. This review discusses the current drawbacks of the diagnostic work-up in iliac artery disease and the potential benefits of targeted iliac artery lesion selection with the aid of more advanced functional imaging. Novel techniques to decrease radiation exposure and the amount of CM are also discussed.

2. Contemporary diagnostics of equivocal iliac artery stenoses

When clinical symptoms justify further evaluation noninvasive imaging methods to screen patients with PAOD are DUS, CTA, and MRA [27–32]. These techniques provide the physician with detailed anatomic information of the vascular tree and the extent and location of the lesions. A major limitation of CTA and MRA is that these techniques only provide an anatomic assessment but offer no hemodynamic information [33]. The hemodynamic significance of a single stenosis or serial stenoses depends on a combination of unique physiological and anatomical characteristics, such as flow (pattern), stenosis length, wall stiffness and collateral vessels, that are different for each patient. DUS can provide hemodynamic information on iliac stenoses by using the peak systolic velocity (PSV) or the PSV ratio (PSVR), but these measurements can only be performed during rest and do not represent the physiological state during maximum hyperemia (exercise) [5]. Furthermore, DUS is operator dependent, can be hampered by gaseous interposition of bowels and it is more difficult to determine optimal C-arm settings for treatment planning.

A widely accepted geometric cutoff value to define a hemodynamically significant stenosis is a 50% lumen diameter reduction. The geometric cutoff value originates from earlier in-vitro and in-vivo studies that have shown that a ‘critical arterial stenosis’ in iliac arteries occurs at 75–80% lumen area reduction or 50% lumen diameter reduction [34–36]. These studies have also shown that beyond these points there is a very rapid decrease in flow and rise in pressure gradient.

From a physiological perspective, a stenosis with a mean arterial pressure (MAP) gradient ≥10 mm Hg under hyperemic conditions is considered hemodynamically significant [37,38]. Only a moderate correlation between 50% lumen diameter reduction and intra-arterial pressure gradients exists [5,39], however, due to the great variance in flow decrease and pressure gradients caused by an equivocal stenosis, only a small change in lumen diameter may lead to a completely different hemodynamic outcome [34–36,40–42]. Determining the exact pressure gradient of equivocal stenoses before a patient is planned for an invasive intervention is therefore worthwhile. Categorizing equivocal stenoses as hemodynamically significant and nonsignificant lesions allows for targeted treatment.

Several studies investigating the performance of a geometry-based diagnostics of a hemodynamically significant stenosis have shown limited diagnostic accuracy. Digital subtraction angiography (eyeballing the stenosis), quantitative vascular analysis, and PSVRs on DUS have a diagnostic accuracy of 71% to 83% [5,39,43] compared with intra-arterial pressure measurements. Therefore, geometric measures alone remain insufficient to determine whether an equivocal stenosis is hemodynamically significant [4,5,44,45].

Intra-arterial translesional pressure measurements (under hyperemic conditions) remain the gold standard for identifying stenoses requiring treatment [46], but these are invasive and require more radiation exposure. Major complication rates of 4.3% to 5.2% in endovascular iliac procedures have been reported [47]. Therefore, optimization or enhancement of judgment on noninvasive imaging methods is required to minimize unnecessary invasive imaging procedures and overtreatment of lesions that are not hemodynamically significant.

3. Advanced diagnosis of suspected iliac artery stenotic disease

Several computational fluid models to calculate translesional pressure gradients on pre-procedural imaging have been proposed for equivocal stenoses in coronary arteries [8–10,48–51]. Computational fluid models are based on principles of fluid mechanics. Usually 3D models are utilized, but these models are time consuming. 2D models only require several minutes to process, but a 2D model assumes that the vessel and stenoses are axisymmetric. However, an in vitro study has shown that an eccentric stenosis leads to a similar pressure gradient as a circular stenosis of the same area and does not influence the outcome of a 2D model [52].

Recently, a patient-specific two-dimensional (2D) computational fluid model for equivocal iliac artery stenoses was assessed [7]. This model was based on 3D rotational angiographic image information and validated with intra-arterial pressure measurements under hyperemic conditions (DETECT-PAD trial; http://www.trialregister.nl; identifier: NTR5085, registered on 9 March 2015). This model enables a prediction of whether the mean arterial pressure gradient across an equivocal stenosis is hemodynamically significant (≥10 mm Hg) or not significant (<10 mm Hg), with sensitivity of 95%, specificity of 60%, and overall predictive value of 88% [7].

A major limitation is that 2D and 3D models cannot provide instant results because they are time consuming, and additional computational fluid dynamics software packages are required to perform the calculations. However, geometric models have recently been developed to estimate the intra-arterial pressure gradient in common and external iliac artery stenoses under hyperemic conditions in seconds and without the need for an additional computational fluid dynamics software package [11,53]. These models were validated with geometric input parameters derived from angiographic images from the DETECT PAD trial. The results of these geometry-based models were comparable to those of the more time-consuming 2D model [11,53]. However, a drawback of these models is that the patient-specific geometric properties are still obtained during catheterization at the angiography suite. The next step is therefore to validate the geometric model for preprocedural MRA or CTA. Another potential drawback of these models is that they do not include collateral circulation which is a major source of error in pressure gradient estimation. However, the models are built to estimate pressure gradients in equivocal stenosis. Collateral formation is almost exclusively seen in chronic total occlusions (CTO) and in very severe stenoses (>75%) and there will be no doubt on preprocedural imaging or angiography whether these lesions are hemodynamically significant making intra-arterial pressure measurements unnecessary. In general, equivocal stenoses have not yet developed collateral circulation.

Another approach to determine hemodynamic significance without the use of intra-arterial pressure measurements is to use machine learning (ML) [54,55], of which the first reports have been published in the cardiac literature. Machine learning is the study of computer algorithms that allows computer programs to automatically improve through experience [56]. To predict whether the fractional flow reserve (FFR) was ≤0.80, Cho and coworkers [55] used a data set of 1501 patients with coronary angiograms and Lee and coworkers [54] used a data set of 1328 patients with intravascular ultrasound images to train their respective models. Both studies achieved an overall accuracy of 82%. A drawback of these models is that both still require an invasive procedure (angiography). To develop ML models for predicting the hemodynamic significance of an iliac artery stenosis, large amounts of clinical data (preprocedural imaging and intra-arterial pressure measurements) need to be available. Obtaining sufficient data will be a major challenge, because performing intra-arterial measurements under hyperemic conditions is not a daily routine for iliac artery disease.

4. Periprocedural visualization and fusion techniques to reduce radiation exposure and iodinated CM

A preprocedural CTA or MRA can help to identify challenging anatomy and determine optimal angulation settings of the C-arm during the procedure. Repeated image acquisition as the C-arm is positioned through trial and error, commonly referred to as ‘fluoro-hunting’, must be avoided because it takes time and increases radiation exposure to the patient and the endovascular team [57]. This is one of the major drawbacks of DUS.

Angulation of the C-arm, which results in more mass that needs to be traversed and more scattered radiation, is especially correlated with a significantly higher radiation dose for the patient and the endovascular team [58,59]. Determining the optimal viewing angles of the aortic and iliac bifurcations on preprocedural imaging can avoid unnecessary radiation exposure.

Fusion imaging was introduced in 2011 [19]. The aim of fusion-image guidance is to facilitate navigation with catheters and guidewires by continuously projecting the vascular tree acquired from preoperative imaging (CTA or MRA) onto live fluoroscopic images, thus reducing the need for iodinated CM injection (Figure 1) and periprocedural acquired high resolution imaging, thus reducing radiation dose.

Figure 1. Example of an iliac image fusion with overlay of a vascular tree.

Image fusion can be performed using a 2D/3D or a 3D/3D registration. A 2D/3D fusion is based on 2 single-shot orthogonal exposures, whereas the 3D/3D fusion uses a noncontrast-enhanced CBCT scan. During the CBCT scan, fluoroscopic images are continuously acquired over an angle of at least 180° and then reconstructed into a 3D data set. 3D/3D registration may lead to better accuracy compared with 2D/3D registration [60], but at the expense of a higher radiation dose [61,62]. To correctly match the preprocedurally and periprocedurally acquired images, landmarks have to be defined. When 2D/3D or 3D/3D registration is performed, these landmarks consist of bony structures such as vertebrae, femoral heads, or the iliac crest, and in case of a preoperative CTA, aortic wall calcifications. Acquisition of these images and registration with preprocedural imaging is usually performed in less than 10 minutes.

A drawback of fusion techniques is that the registration is rigid, whereas the position of the patient often changes during the procedure compared with positioning during the preprocedural imaging. Also, the introduction of delivery devices and stiff guidewires can challenge the use of image fusion. Deformation and displacement of the iliac arteries will occur and cannot be corrected with a rigid fusion technique.

To overcome these drawbacks of rigid fusion, several solutions regarding distortion correction, automatic segmentation and registration have been proposed [63–67]. To our knowledge the only commercially available system is a cloud-based artificial intelligence (AI)-assisted 3D fusion technique that has been developed by CYDAR Medical (Barrington, United Kingdom). With this AI-assisted nonrigid registration, the vertebral anatomy on fluoroscopy and the preprocedural imaging are compared. An adjusted 3D vascular tree is then overlaid on the fluoroscopy screen, which will continuously update and verify itself [65]. This does not only allow for a translation of a rigid mask, but a deformation of the mask as well. It has been shown that this system could also correct for displacements introduced by stiff devices [66]. The efficacy and added value of the novel technique has only been evaluated in a retrospective study [67].

Another periprocedural imaging technique is intravascular ultrasound (IVUS). IVUS can be very useful for detecting true lumen size, evaluating residual stenosis after plain old balloon angioplasty and detecting technical failure (eg incomplete stent apposition, thrombosis, residual dissection) [68]. Although IVUS can provide a high-resolution cross-sectional image of the lesion [69], it remains a geometric measure and cannot provide hemodynamic information. The high costs of IVUS, the extra periprocedural actions that need to be taken and the increasing quality of MRA and CTA is probably why IVUS for iliac procedures has not been incorporated into the guidelines or clinical practice.

4.1. Image fusion in obstructive iliac disease; 3DMR-Iliac-Roadmapping study

The 3DMR-Iliac-Roadmapping study (https://www.trialregister.nl/; identifier NTR5008; registered on 16 December 2014) was the first prospective, multicenter, randomized trial designed to evaluate whether 3D image fusion during endovascular iliac artery interventions could reduce the amount of administered iodinated CM. The study protocol, which was approved by the Maastricht University Medical Center+ and Maastricht University ethics committee (METC azM/UM), was previously published [70], and all study procedures were conducted in accordance with good clinical practices and applicable laws.

The 3DMR-Iliac-Roadmapping study included 41 patients; of these, 39 were randomized, and 2 were not randomized due to logistic issues. Patients were randomized into 2 groups: treatment with 3D/3D image fusion (n = 19) or without image fusion (n = 20). Demographics, comorbidities, and cardiovascular risk factors were similar between the 2 groups (Table 1). Impaired renal function, defined as estimated glomerular filtration rate (eGFR) ≤60 mL/min/1.73 m², and bodyweight (≤90 or >90 kg) were used as stratification factors. Each group had 3 patients with an eGFR ≤60 mL/min/1.73 m² and 3 patients with a body weight >90 kg. After inclusion of 40 patients, as defined in the study protocol, an interim analysis was performed by an independent safety committee.

Table 1. Baseline characteristics of the 3DMR-Iliac-Roadmapping study population^a,b,c.

				group with fusion (n = 19)		Group without fusion (n = 20)		P
Age (years)				67.0	[50–88]	66.5	[50–86]	0.767 ^c
Sex			Male	13/19		15/20		0.731 ^b
Rutherford categories
	Rutherford category 1			1/19 5/19		1/20		1.000 ^b
	Rutherford category 2			5/19		6/20		1.000 ^b
	Rutherford category 3			9/19		9/20		1.000 ^b
	Rutherford category 4			3/19		4/20		0.106 ^b
	Rutherford category 5			1/19		0/20		0.342 ^b
Functional parameters
		ABI in rest		0.6	[0.4–0.9]	0.6	[0.3–0.9]	0.55 ^c
		ABI after exercise		0.2	[0.1–0.8]	0.3	[0.1–0.8]	0.94 ^c
		Pain-free walking distance	meter	85	[10–300]	100	[30–2000]	0.44 ^c
Creatinine			µmol/l	81	[59–247]	90	[45–120]	0.51 ^c
Risk factors
	Diabetes		Yes	5/19		3/20		0.45 ^b
	Smoking, current or recent		Yes	13/19		13/20		1.00 ^b
	BMI		Kg/m2	26.5	[18.6–34.7]	24.4	[18.3–34.9]	0.24 ^c

ABI, ankle brachial index; BMI, Body Mass Index.

^aContinuous data are shown as median [range]

^bPierson’s Chi-square and Fisher’s Exact test where appropriate, unless otherwise indicated

^cOne-tailed Mann-Whitney U test

Total median volume of CM used was 32 mL (range, 7–112 mL) in the group with fusion and 42 mL (range, 20–197 mL) in the group without fusion (P = 0.254). All endpoints are listed in Table 2. Technical success, defined as the ability to successfully perform plain balloon angioplasty or stent placement with a residual stenosis <30%, was achieved in treated patients. 2 patients in the group without fusion were not treated; in 1 patient it was decided after DSA to replan the patient for a hybrid approach, in the other patient it was concluded the stenosis was non-significant. Similarly, fluoroscopy time, procedure time, and radiation dose did not differ significantly between the groups. Fusion accuracy registration in the fusion group (n = 19) was scored in 6 patients as accurate (<2 mm difference), in 10 patients as mismatch (2–5 mm difference), in 2 patients as inaccurate (>5 mm difference) and was missing for 1 patient.

Table 2. Study end points of the 3DMR-iliac-roadmapping study population^a,b,c.

		Group with fusion (n = 19)		Group without fusion (n = 20)		P
Primary endpoint
Volume iodinated contrast	ml	32	[7–112]	42	[20–137]	0.25 ^c
Secondary endpoints
Technical success (<30% residual stenosis)	Yes	19		18		0.49 ^b
Fluoroscopy time	mm:ss	05:27	[01:29–32:25]	04:12	[01:11–45:48]	0.93 ^c
Procedure time	minutes	46	[13–111]	37	[13–124]	0.93 ^c
Radiation dose (DAP)	mGycm2	48,955	[1430–15,900]	23,986	[1717–95,782]	0.96 ^c
Air Kerma (AK)	mGy	188	[71–754]	144	[25–3355]	0.83 ^c

DAP, dose area product; AK, air kerma

^aContinuous data are shown as median [range]

^bPierson’s Chi-square and Fisher’s Exact test where appropriate, unless otherwise indicated

^cOne-tailed Mann-Whitney U test

The primary end point showed no significant decrease in the total volume of iodinated CM used in the fusion group at the mandatory interim analysis. The committee reported that even if statistical significance might be reached, the amount of iodinated CM reduction was unlikely to lead to a clinically relevant reduction in CM. Based on the estimate, the study was terminated prematurely.

4.2. Dilution of iodinated CM

The use of iodinated CM should always be prospectively monitored and evaluated in order to facilitate improvement and a potential reduction in volumes used. Current guidelines recommend the use of either low-osmolar contrast media (LOCM) or iso-osmolar contrast media (IOCM) for endovascular use since there is no difference in incidence of post contrast acute kidney injury, acute reactions or acute adverse events [15,71].

An alternative to reduce the total CM volume is to dilute the CM with saline. With the current standard of digital flat-panel detectors, dilution of CM should be considered and is possible without losing diagnostic information or image quality. Unfortunately data to justify this claim is scarce. A randomized controlled trial showed non-inferiority for using CM with 140 mgI/mL instead of 300 mgI/mL for infrainguinal angioplasty [72].

Furthermore, the use of a single-head injector can reduce the total CM volume when compared to manual injections, since they are more precise. With dual-head injectors it is possible to adjust CM dilution to patient and procedural characteristics (e.g. obese patient or a lateral view) by adjusting the dilution per injection.

4.3. Device visualization and tracking

Fluoroscopy has traditionally been used to visualize guidewires, catheters, and other endovascular devices. With image fusion guidance, it is possible to decrease catheterization time, leading to a shorter fluoroscopy time and reduced radiation dose [73]. The next step would be to track catheters and guidewires without the need for fluoroscopy.

A not yet commercially available and new visualization technique, called Fiber Optic RealShape (FORS), was recently introduced by Philips Medical Systems Nederland (Best, The Netherlands). By incorporating hair-thin optical fibers into catheters and guidewires, the shape of these endovascular devices can be visualized using light pulse reflection (Figure 2). This provides a real-time 3D visualization of the shape of the device, which can be projected on a preprocedural or periprocedural acquired image. The image volume needs to be registered to the patient using a conventional 2D/3D or 3D/3D X-ray registration, which is performed using dedicated software for the FORS technology [74]. When the image volume is aligned with the patient, the FORS enabled devices need to be registered. This is done similarly to the image volume registration using a 2D/3D registration.

Figure 2. Incorporation of hair-thin optical fibers into a catheter (courtesy of Philips Medical Systems Nederland, Best, The Netherlands).

The possible major advantage of this novel technique is that it provides real-time 3D feedback on the position of the guidewire or catheter. The endovascular specialist will be able to navigate a guidewire or catheter without the use of fluoroscopy in combination with 3D information, resulting in a reduction of radiation exposure and iodinated CM use. The 3D projection capabilities also exceed those of conventional 2D fluoroscopy in a complex and elongated anatomy, because the endovascular specialist can use different views of the tracked device simultaneously. Current drawbacks are that only 2 devices can be tracked at the same time, and only FORS enabled devices can be tracked using the proprietary hardware and software.

Preclinical and first-in-man feasibility results were presented at Aortic Live 2018 and at the Leipzig Interventional Course (LINC) 2019 and 2020 and showed promising results regarding several navigation tasks [23–25]. Clinical trials are expected to start shortly.

Another new kid on the block uses electromagnetic fields to track catheters and is termed electromagnetic tracking. Tracking is performed by using a field generator that creates a magnetic field of known geometry and sensors that measure magnetic flux or magnetic fields. Because the geometry of the emitting coil and the current through the coil are known, the shape and properties of the reference field can be calculated using the law of Biot-Savart [26]. This allows for the calculation of the field strength at various points. By incorporating a magnetic sensor into a medical device, the exact location of the sensor can be calculated in relation to the reference field, and movement tracking is possible with 6 degrees of freedom (along the x, y, and z axis and rotation on each axis).

Registration of preprocedural images can be performed in various ways, including 2D/3D, 3D/3D, with the use of external fiducials, and more recently an original registration method has been proposed using only catheters with incorporated electromagnetic sensors [75]. With this method, a catheter pull-back is performed in the iliac axis. The recorded path and the preprocedural images can be used for fusion registration. The recorded mean registration error in a phantom setup was 1.3 mm (range, 0.88–1.89 mm) [75]. A potential drawback of electromagnetic tracking is that only the sensor, and not the whole catheter, can be tracked. Catheters can be projected onto the screen using its known dimensions in relation to the sensor, but bending or curves of the catheter are not tracked. Several preclinical animal and phantom studies have been published [75–79], but to our knowledge, no in human studies have been published so far for endovascular indications.

Endovascular specialists have familiarized themselves with the different properties of various guidewires and catheters, and their behavior in different situations regarding steerability and push. Companies introducing new, trackable devices will be challenged to not sacrifice the abilities endovascular specialists are used to work with. Open vessel navigation will most likely not be a problem, but it remains questionable how these new tools will behave in challenging situations like long CTOs.

5. Conclusion

Targeted treatment for the individual patient by using a combination of predictive modeling, image fusion, and device visualization or device tracking will play a major role in the next years of iliac endovascular procedures. It may allow endovascular specialists to reduce unnecessary, invasive procedures. Owing to the already low CM volumes used in iliac procedures, image fusion alone has no relevant impact on the reduction of the use of iodinated CM in straightforward iliac procedures.

6. Expert opinion

An endovascular approach has become the preferred treatment for iliac artery disease, with good technical and long-term outcomes. The number of procedures for iliac artery stenotic disease is gradually increasing each year [80] and the need to improve and develop new endovascular (imaging) techniques to minimize the risks for patients and endovascular specialists is important.

In the era of personalized medicine, the preprocedural workup for endovascular iliac procedures should incorporate a targeted and patient-specific treatment plan. Currently, optimal treatment planning in case of equivocal stenoses should include intra-arterial pressure measurements, but this technique is not commonly used [81]. The dominance in clinicians’ decisions by visual estimation, weakly correlated to intra-arterial pressure gradients, may lead to a higher rate of unnecessary iliac artery interventions. By using predictive modeling, either computational flow modeling or machine learning, lesion selection can be performed on preprocedural imaging. However, the question of which and when a model is judged good enough to be implemented in daily practice remains.

Fryback and Thornbury describe a hierarchical model of efficacy for diagnostic imaging. The model outlines the optimal level of evidence necessary to prove the value of a diagnostic imaging test. It addresses factors related to technical quality and diagnostic accuracy, diagnostic and therapeutic impact, and patient and societal outcomes [82]. Applying this to the computational model as described in the DETECT-PAD trial, technical efficacy has been proven, but much more clinical data, ranging up to 200 to 300 patients, will have to be acquired from new studies.

Radiation exposure during endovascular procedures is an invisible threat, and attention to protection can be easily forgotten, especially in complex procedures. Body mass index and lesion complexity are associated with significantly higher radiation exposures during the procedure [83,84]. With an increase in procedure numbers, an increase in patient obesity and multimorbidity, and more complex lesions, the cumulative lifetime dose of the endovascular staff will increase. The ALARA (as low as reasonably achievable) principle is important to maintain. For the first time, radiation has been shown to cause DNA damage in endovascular specialists [85]. However, in light of the growing evidence that the use of image fusion can reduce the radiation dose for both the patient and endovascular team, the observation that image fusion has not been adapted in international scientific guidelines or clinical workflows is remarkable but we expect that this will be amended in coming updates of various international guidelines.

Workflow issues often arise with the introduction of new technology. When radiation reduction technology is not directly in line with the expectations of the endovascular team, or using a new technology is cumbersome, adaptation will be less likely [86] and may even lead to neglect of the harmful effects to both the patient and staff. Rigorous and continuous training is required to safely and effectively introduce a new workflow or technology.

The 3DMR-Iliac-Roadmapping study is the first multicenter randomized controlled trial of image fusion compared with regular imaging in an endovascular iliac artery procedure. The results show that image fusion in endovascular iliac artery procedures is feasible and safe. With image fusion the main focus has been to fuse preprocedural CT images. The 3DMR-Iliac-Roadmapping study shows it is feasible to use preprocedural MRA for image fusion. No other iliac fusion studies have been published, and the CYDAR Iliac trial, focused on patient and staff safety regarding radiation dose using CYDAR AI, was expected to complete inclusion in January 2020 (https://clinicaltrials.gov/; identifier NCT03713450).

The 3DMR-Iliac-Roadmapping study showed a median difference in the administered volume of iodinated CM of 10 mL, but how much CM reduction justifies a complete overhaul of clinical workflow? Contrast-induced nephropathy (CIN) is dose-related to the amount of iodinated CM in very high-risk patients [87], and multiple efforts have been done to determine safe procedural doses [88–91]. Yoon and Hurr suggested the use of grams of iodine to the eGFR ratio and analyzed that a ratio of 1.42 g iodine/eGFR had a sensitivity and specificity of 81.3% and 80% for developing CIN [91]. The difference in grams of iodine in our study was only 3.0 g (based on median administered volume of CM and a concentration of 300 mg/mL), and this is unlikely to have a relevant clinical effect, even in high-risk patients. This suggests that image fusion in straightforward iliac procedures has no relevant impact on the reduction of the use of iodinated CM.

We believe that the current developments regarding predictive modeling and catheter visualization or tracking, in combination with image fusion, are very promising and will change the current endovascular field in the next 5 to 10 years. Endovascular specialists will only use augmented diagnostics tools, and eyeballing may become obsolete. Imaging systems will be able to determine a hemodynamic gradient instantaneously and include it in a report. Also, to a certain degree, it should be technically feasible to perform endovascular procedures without the need for CM or radiation like with the FORS technology. Failure rates for complex lesions like CTOs will become less. Often these lesions need to be transversed in the subintimal plane and reentering the true lumen can be challenging. By being able to track the wire position relative to the true lumen, determining an optimal reentry angle will become easier. We do believe that fluoroscopy guided interventions will stay around for a long time to come, but the need for endovascular specialists to solely rely on fluoroscopy will become less and less.

A big challenge however, will be the lack of reimbursement or insufficient reimbursement to compensate for the introduction of new devices and the early adaptation to new technology and workflows. Generally new technologies come at a premium price, while often initially lacking sufficient data that demonstrates better patient outcomes. Often medical doctors are dependent on medical device companies for the (co-)development of new technologies like CYDAR or FORS. We would like to challenge medical device companies to provide more quality and efficacy data upon market introduction while maintaining an affordable price level. This data is needed to support a systematic evaluation process and informed decision-making by healthcare providers. Only then can healthcare providers invest into new technologies in a financially sustainable way while having the biggest impact on patient outcome.

Article highlights

• In the era of personalized medicine, the preprocedural workup for endovascular iliac procedures should incorporate a targeted and patient-specific treatment plan.

• Intra-arterial pressure measurements remain the golden standard for equivocal iliac stenoses, but in the near future lesion selection could very well be performed on preprocedural imaging using predictive modeling.

• Prospective monitoring of iodinated contrast media usage and dilution should be considered to improve injection protocols and decrease the use of contrast media.

• The use of preprocedurally acquired MRA images for image fusion is feasible and can be considered instead of CTA images.

• Catheter visualization and tracking techniques, in combination with image fusion, will change the current endovascular field in the next 5 to 10 years.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Declaration of interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Additional information

Funding

This paper was not funded.