Intravascular Lithotripsy-Assisted Transfemoral Transcatheter Aortic Valve Implantation After Failed Balloon Angioplasty in Patients With Severe Calcified Peripheral Artery Disease

© 2024 HMP Global. All Rights Reserved.
Any views and opinions expressed are those of the author(s) and/or participants and do not necessarily reflect the views, policy, or position of the Journal of Invasive Cardiology or HMP Global, their employees, and affiliates.

J INVASIVE CARDIOL 2024. doi:10.25270/jic/24.00049. Epub July 18, 2024.

Objectives. Calcific peripheral artery disease (PAD) is a common finding in patients scheduled for transcatheter aortic valve implantation (TAVI) and often requires iliofemoral axis preparation. However, evidence about the use of intravascular lithotripsy (IVL) in this setting is scarce. The aim of this study was to describe in-hospital and mid-term outcomes of IVL-assisted transfemoral (TF)-TAVI in patients with severe calcific PAD.

Methods. The study included 13 patients with severe calcified PAD who were initially scheduled for balloon angioplasty (PTA)-assisted TF-TAVI but were eventually treated with peripheral IVL between October 2020 and February 2024. Accurate analysis of preoperative computed tomography scans revealed difficult TF access routes (severe calcified PAD with minimal lumen diameter ≤ 4.5 mm, circumferential calcification along iliofemoral axis, and marked vessels tortuosity).

Results. In all cases, IVL was performed after PTA failure and allowed successful valve delivery. One patient had procedural bleeding (BARC-3a). The valve was successfully delivered without complications in 12 patients (92.3%), and no major adverse events were reported at the longest follow-up (median 18.5 months).

Conclusions. IVL-assisted TF-TAVI may represent a feasible and safe option for patients presenting with severe aortic stenosis and severe calcified PAD. However, standardization of the access site severity as well as indications for a planned up-front IVL-assisted strategy are missing and require dedicated studies.

The transfemoral (TF) approach is currently the preferred route to perform transcatheter aortic valve implantation (TAVI) because it provides better outcomes than non-TF approaches.^1-3 However, vascular complications still occur in 9% of TF-TAVI cases, even in high-volume centers and with latest valve models. Furthermore, in approximately 5% to 15% of cases, the presence of severe peripheral artery disease (PAD) may compromise the feasibility and the safety of a TF approach, requiring iliofemoral axis preparation through balloon angioplasty (PTA).⁴ Particularly, extensive calcifications and tortuosity in vessels of small diameter are associated with poor prognosis and increased cardiovascular mortality and morbidity.⁵ Proficiency with peripheral PTA and the recent availability of peripheral intravascular lithotripsy (IVL) may expand access to TF-TAVI.⁶ While recent studies have proved that IVL adds safety to the treatment of calcific PAD,^7-9 the role of IVL in TF-TAVI is not yet defined.

The study by Nardi et al on hostile TF vascular access¹⁰ provides an important starting point to investigate the potentials of such a new therapy in this challenging setting. It is intuitive that highly calcific and severely tortuous vessels with narrowed lumens are expected to be troublesome during the valve delivery, hence are considered “hostile”, but beyond this qualitative assessment derived from subjective evaluation of the iliofemoral anatomy at the preoperative computed tomography (CT) scan, no grading is available to quantify the vascular access hostility in order to identify levels of risk or contraindications to a TF attempt.¹¹ Likewise, there is no guide for preparing vascular access according to its “hostility” (PTA, IVL or surgery). Preoperative CT scan is the gold standard for assessing the complexity of iliofemoral anatomy as it provides an accurate identification and measurement of the minimal lumen diameter (MLD), the characterization of calcifications along the vessels (both in terms of circumferential and longitudinal distribution), and, lastly, the assessment of the axis tortuosity. However, threshold values discriminating between complex and simple anatomies are lacking, which makes the standardization and objective classification of access hostility impossible.¹²

Herein, we report a case series of consecutive TAVI procedures that were considered unsuitable for the TF approach by the heart team after PTA failure and were consequently performed with IVL assistance in a center with documented proficiency in peripheral PTA.⁴ The aim was to describe the rate of valve delivery and procedural outcomes in-hospital and at mid-term follow-up.

Clinical and radiological criteria for patients’ selection. Patients included in this case series (n = 13) were pooled from the Verona Valve Registry, a prospective study approved by the local ethical committee (CESC, n = 1918), between October 2020 and February 2024. The included patients had severe calcified bilateral PAD as revealed from the CT pre-TAVI scan and were initially planned to undergo upfront PTA-assisted TF-TAVI, but eventually required additional vessel modification with IVL to achieve successful valve delivery. The study was conducted according to the guidelines of the Declaration of Helsinki and was approved by the Institutional Review Board of University of Verona. Informed consent was obtained by all subjects included in the study.

Severe calcified PAD was defined as a minimal lumen diameter (MLD) less than or equal to 4.5 mm combined with circumferential calcification (ie, 180º calcification at the vessel cross-section of the tightest site) in vessels with an external diameter of at least 5 mm and having marked tortuosity (at least 1 angulation ≥ 50°). Patients who had undergone previous vascular interventions (either surgical or endovascular) at the iliofemoral access chosen for TAVI were not included in the series.

Preoperative CT scans were analyzed with a dedicated software (3Mensio, Structural Heart, Pie Medical Imaging) by trained analysts using the specific femoral access analysis package included with the software. All the included analysis were performed on contrast-enhanced CT scans.

MLD was first measured in both the left and right iliofemoral axes, and the largest vessel was chosen as the potential operative access. Patients were included if they presented a minimum Reference Vessel Diameter (RefD) of 5 mm, measured proximally to the tightest segment (where MLD was measured) in a healthy segment. After 3-dimensional (3D) reconstruction of the chosen axis, the analyst systematically performed the following measurements:

1. Circumferential distribution of calcifications at the tightest site: once the MLD spot was identified, the distribution of the calcium along the cross-section circumference was graded according to degrees of its extension. Circumferential calcification was defined when at least 180º (Figure 1)
2. Longitudinal distribution of calcifications: from the “stretched-vessel” view, using the filter for calcifications, the maximum length of the calcified wall across the entire diseased segment was measured as the percentage of calcium length over the vessel length (defined as diffuse calcifications when > 50%) (Figure 2).
3. Vessel tortuosity: exploiting the “tangent angle” function, degrees of central line angulation proximal and distal to the MLD point were measured, obtaining a grade of the vessel curvature at the site of the MLD. Tortuosity was defined in the presence of at least 50º between the proximal and distal segments angles compared to the MLD site (Figure 2).
4.  

Notably, the afore-described parameters offered the advantages of being objective and reproducible as they were semi-automatically computed by the software.

Description of the technique. In all patients, femoral puncture was performed through ultrasound guidance following local anesthesia administration of lidocaine. Conscious sedation and contralateral femoral access (6-French [Fr] sheath) were used routinely. Closure of the puncture site was performed with the pre-implanted Perclose ProGlide suture-mediated closure system (Abbott).

The valve type (Sapien 3 or ULTRA [Edwards] or CoreValve Evolut-R or Pro [Medtronic]) and size were chosen according to the anatomical characteristics and valve morphology obtained from the pre-procedural CT-scan analysis.

After access puncture and wiring of the operative access with a 0.035-inch wire, PTA was initially attempted at the tightest sites in all cases with non-compliant balloons; however, successful valve advancement was not achieved. Therefore, additional dilatations with the IVL balloon were performed using the Shockwave Peripheral IVL catheter (Shockwave Medical). IVL therapy was delivered in 30 pulses per cycle. Each cycle could be repeated, as needed, until satisfactory luminal diameter gain was obtained and either the valve or the sheath could gain the abdominal aorta. Additional PTA with a non-compliant balloon was performed, as needed, to further increase luminal gain following calcium fracture. Figure 3 provides a step-by-step description of the IVL technique.

Procedural details and clinical outcomes of the included patients. The baseline characteristics are reported in Table 1. All patients underwent TAVI electively and under stable hemodynamic conditions.

At the CT, the mean MLD was 4.4 ± 0.2 mm, with 4 cases presenting 1 spot with MLD less than or equal to 4.0 mm, and 2 cases presenting 2 sites less than or equal to 4.0 mm along the same iliofemoral axis (Table 2).

All cases presented at least a circumferential (≥ 180º) spot of calcium in correspondence of the MLD, while most of the cases presented diffuse calcifications spanning across the entire length of iliofemoral axis irrespective of the location of the narrowest spot (≥ 50% of diseased length of the vessel) (Figure 1B).

A tortuous segment at the level of MLD was identified in all cases, with 1 case (patient 11) also presenting a severely protruding speckle of calcium (≥ 50% of the lumen area) at the level of right common iliac artery. (Figure 4).

A balloon-expandable valve (Sapien 3/ULTRA) was implanted in 9 cases (69.2%), while a self-expandable valve (Evolut PRO/PRO+) was chosen for the remaining 4 cases (30.8%). The size of the delivery systems was 14 Fr in 11 cases (84.6%) and 16 Fr in 2 cases (15.4%). 14-Fr InLine (Medtronic) sheaths were used in the 4 self-expandable valve cases (30.8%), while eSheaths (Edwards) were used in the 9 remaining cases (69.2%; 1 used 18 Fr, 2 used 16 Fr, and 6 used 14 Fr).

Right transfemoral access was used in 10 cases (77%) and left femoral in 3 (23%), according to the larger MLD as measured by CT analysis. In 4 cases (30.8%) with severe vessel tortuosity and presence of bulky and asymmetric calcifications confirmed at IVUS, PTA was initially performed with a smaller non-compliant balloon, compared to reference vessel diameter. The main reason for the choice of a less than 1:1 balloon to vessel ratio in such cases was the operator’s concern about vascular rupture.

Conversely, the remaining 9 cases (69.2%) were initially approached with a 1:1 sized non-compliant balloon, either measured at preoperative CT scan or reassessed during the procedure with intravascular ultrasound (IVUS). No oversizing (> 1:1 balloon to vessel ratio) was performed in any case.

Each patient underwent 4 to 6 inflations using non-compliant balloons with an average diameter and length of 6.4 ± 1.6 mm and 62 ± 14 mm, respectively. After initial PTA, IVUS re-check was used in 8 cases (61.5%), guiding the step-up to IVL when balloon waist was still present despite multiple inflations. Further attempts to deliver the valve immediately after PTA were not performed for these 8 patients, as IVUS findings already confirmed insufficient plaque modification. In the remaining 5 cases (38.5%) balloon inflation appeared to be satisfactory at the fluoroscopy; thus, delivery of the valve was unsuccessfully re-attempted immediately after PTA but before IVL.

The mean number of delivered IVL cycles (30 pulses/cycle) was 3.1 ± 0.9 for each patient. The mean diameter and length of the used IVL balloons were 6.2 ± 1.6 mm and 60 ± 0 mm, respectively, with an average maximal pressure inflation of 6.4 ± 0.8 atmosphere.

In 9 cases (69.2%), the valve was successfully advanced immediately after IVL performance without the need of further balloon dilatation. In the remaining 4 cases (30.8%), additional inflations with non-compliant balloons were performed because of un-satisfactory lumen gain at either post-IVL angiogram or post-IVL IVUS.

Examples of a sub-optimal result and a good result are displayed in Figure 5. None of the patients ended up receiving or requiring any stent placement along the iliofemoral axis. Overall, the mean procedural time was 134 ± 83 minutes, the mean fluoroscopy time was 37 ± 13 minutes, and the mean contrast volume was 125 ± 58 mL.

All patients achieved successful femoral closure with the pre-implanted ProGlide. Four patients (30.8%; Patients 5, 7, 12, and 13) experienced temporary thrombotic occlusion of the ipsilateral superficial femoral artery after treatment of the stenotic segment with an IVL balloon despite adequate heparinization as tested through activated partial thromboplastin time. These vascular complications were promptly and effectively resolved with adjunctive and selective heparin injection in 2 patients and with prolonged balloon inflation in the other 2 patients.

Final iliofemoral angiography in these 4 patients showed the absence of luminal obstruction and good blood flow without the need for further intervention. A major bleeding (diagnosed with lab tests showing a significant hemoglobin [Hb] drop of 3.2 g/dL) was reported for Patient 11 following a particularly complex and prolonged procedure (procedural time 201 min; fluoroscopy time 52 min; contrast volume 280 mL). In this case, Hb stabilized after the first drop (lowest Hb: 7.6 g/dL) and required a transfusion of 3 units of red blood cells. A clear source of bleeding was not detected. The doppler ultrasound of the vascular accesses did not show complications. Likely, the Hb drop was due to the numerous maneuvers required for the vessel preparation and the multiple attempts to advance the delivery system, leading to a non-overt but constant blood loss from the iliofemoral access site.

Overall, the mean length of in-hospital stay was 4.8 ± 4.3 days. All patients were discharged home. None of the patients reported significant renal function deterioration, though 1 patient required a permanent pacemaker for atrioventricular block development on postoperative day 2.

All patients were actively mobilized 24 to 48 hours after the procedure. Moreover, lower limb ultrasound doppler control of Patients 5 and 7 confirmed the patency of the operative femoral artery.

None of the patients experienced acute limb ischemia or neurological disorders.

Accordingly, we reported successful valve delivery in all patients (100%) and intra-procedural bleeding in 1 patient (8%); therefore, procedural success without any complication was reported in 12 of the 13 cases (92.3%). Major adverse cardiac and cerebrovascular events at long-term follow-up occurred in 2 patients (15.4%) who died from non-cardiac causes (Patient 2 died after 8 months from COVID-related pneumonia, and Patient 8 died after 11 months from cancer). Longer term follow-up is reported in Table 3, and the definitions of adverse events reported in this series are detailed in Table 4. The requirement for permanent pace-maker implantation was not considered to be a procedural complication in this case series since it was unrelated to the degree of access hostility.

Bearing in mind the limitations due to the small sample size, the presented series suggests the following: (a) IVL-assisted TF-TAVI seems feasible and might be an effective solution in patients presenting objective CT findings characterizing hostile accesses, allowing successful delivery of the valve after ineffective balloon PTA; (b) no IVL-related or major vascular complications were reported in this series, although some access bleeding might derive from the multiple maneuvers required when dealing with most complex cases; and (c) in all, cases, CT scan analysis demonstrated distinctive characteristics of the iliofemoral access, such as the tightest spots with specific calcium distribution combined with tortuosity, indicating challenging anatomy.

At the time of this study, only Nardi et al had reported a relatively large observational experience of 108 patients undergoing TF-TAVI facilitated by IVL.¹⁰ In that study, patients presented calcific PAD characterized by narrowed iliofemoral vessels (mean MLD 4.6 ± 0.9 mm) and extensive circumferential calcifications (average 317.5º). PAD assessment was performed through preoperative CT scan and the use of IVL was based on the operator’s decision as a first strategy and not after failure of PTA attempts. The rate of successful valve delivery was 100.0% with a 3.7% rate of major vascular complications and a 12.0% rate of access-site related complications, mostly due to closure device failure.

Considering that only 1 patient in our case series experienced an in-hospital complication (bleeding event) and there was no need for delaying active patient mobilization or prolonging hospital stay, our experience corroborates the findings observed by Nardi et al in patients in whom the valve could not be implanted despite several attempts of balloon-based PTA.

However, a recent publication of the “Hostile Registry” suggests that the TF route should be avoided in the presence of severely diseased iliofemoral axes because of an increased risk for cerebrovascular events.⁵ In fact, a severe burden of disease detected at the iliofemoral axis may be interpreted as a spy for the presence of further atherosclerotic disease along the entire vascular tree, particularly the aorta. Thus, according to the authors’ interpretation, patients with severe PAD are more likely to develop cerebrovascular events during TAVI.

Although current evidence supports TF-TAVI over surgical aortic valve replacement in elderly patients, outcomes of patients with severe PAD undergoing TF-TAVI are not yet available and, therefore, no clear indications are disclosed in the last guidelines.¹³

Results from the DISRUPT-PAD trial series demonstrate the low incidence of IVL complications (namely post-inflation intimal dissection and subsequent stent implantation) compared with PTA of the femoral arteries alone. In terms of effectiveness, namely the degree of residual stenosis, PTA showed worse results, which supports IVL as a better strategy for preparing the lesion compared with PTA only.^7-9 Likewise, Nardi et el showed that IVL was safe and effective in facilitating TF-TAVI in patients with calcific PAD.¹⁰

Various devices, including orbital atherectomy (OA) devices (eg, Diamondback 36 Peripheral Orbital Atherectomy System, Abbott), directional atherectomy (DA) devices (eg, TurboHawk Peripheral Plaque Excision System and SilverHawk Peripheral Plaque Excision System, Medtronic), and rotational atherectomy (RA) devices (eg, Jetstream Atherectomy, Boston Scientific) have been developed to treat extremely calcific peripheral lesions. OA, characterized by a 360° rotational device with a diamond-coated crown, has proven its efficacy in both coronary and peripheral vessels.  Plaque is removed by the orbital movement of the crown, while the de-bulking area increases with the increase of the rotational speed of the crown. Notably, it is the only atherectomy system that is compatible with 4-Fr sheaths. Giannopuloulos et al highlighted its efficacy alongside balloon angioplasty in the LIBERTY 360 trial, revealing positive outcomes with low complication rates at short-term and reassuring long-term results at 3 years in the context of PAD.¹⁴

Similarly, DA devices demonstrated safety and efficacy in treating claudication or critical limb ischemia at 12-month follow-up in the DEFINITIVE LE study.¹⁵ In DA, plaque is removed by guiding the cutting device of the catheter directly to the plaque, which makes it suitable for the treatment of eccentric lesions. Removed plaque is packed into the nosecone, and, after few passes, the catheter must be retrieved and the nosecone must be emptied in order to proceed with further de-bulking. Thus, a reasonable de-bulking volume is usually achieved after several passes of the catheter, which may lead to increase procedural time and radiation dose.

Jetstream atherectomy catheter is a cutting RA device with active debris aspiration, indicated for both acute thrombus removal and atherectomy of chronic lesions thrombus removal.¹⁶ Micro-and macro-embolization tends to be more likely with such device compared with other de-bulking techniques, so filter protection is advisable.¹⁷

Despite the favorable results in treating PAD, no studies directly compare nor solely describe the aforementioned devices in terms of safety and efficacy in the context of TF-TAVI.

IVL remains the most extensively described device to-date, in the context of vessel wall modification to facilitate the TAVI procedure.

Notably, no head-to-head comparison with conventional balloon PTA has yet been made. A key issue for such a study design is the pre-procedural quantification of the degree of access “hostility”, a concept that lacks standardization and depends on the team experience and discretion.

Palmerini et al recently proposed a score to determine access hostility based on CT scan, however, this focused more on the clinical prognostic comparison (ie, stroke risk) between different accesses (TF vs non-TF) rather than on the identification of variables that may sort out patients who should be excluded from a TF approach or those who require aggressive route preparation to achieve successful TF-TAVI.⁵

Indeed, the definition of hostile vascular access according to specific vessel characteristics before the procedure would certainly provide useful information for the heart team when evaluating the feasibility and the inherent risk of a TF-TAVI in patients with severe PAD. With predictable and reproducible information, vascular access site preparation could be performed accordingly by optimizing resources and containing complications like vessel damage or cerebrovascular events related to complex and prolonged interventional maneuvering. Such information may be also useful for other endovascular procedures that require large bore access, like thoracic and abdominal aneurysm repair.

Modern machinery and advanced software are currently available for both image acquisition and post-processing analysis. A 3D-rendering reconstruction of the entire vascular tree would potentially allow a comprehensive characterization of PAD and its distribution along the vessels. The standardization of adverse variables like small MLD, calcific burden (in terms of calcium arch and longitudinal extension, namely focal versus diffuse), path tortuosity, intra-luminal thrombosis, and puncture site depth are just some examples of the parameters that can be extracted and precisely analyzed during the routine work-up for TAVI.^18-20

Integration of these data could support better identification of candidates for PTA alone or IVL-assisted TAVI and ideally would minimize the risks and maximize the benefits of both strategies. Our analysis suggests that such a population represents about 9% of the patients referred for TAVI in a high-volume referral center.

Within this scope, CT-based risk stratification models should first be developed from retrospective cohorts of subjects like those included in this preliminary analysis. Ultimately, the predictive role of such models should be confirmed through prospective studies directly comparing patients with different degrees of “access hostility”.

Limitations. The single-center experience and the complexity of the PAD in the included patients resulted in the small sample size of the present series. Also, the analyzed CT scan variables have a speculative meaning, and only by expanding the study to a much larger series may the role of the above-mentioned measurements be confirmed. Moreover, despite the user-friendly IVL platform, the presented results derive from the documented experience of the operators involved in peripheral interventions and may vary according to the team’s experience and center volumes. 

Based on the results of this study, IVL preparation of the femoral route seems safe and effective, allowing TF-TAVI deployment even in severely calcified vessels where balloon-PTA assisted vessel preparation failed to permit transcatheter valve navigation to the heart. The integrated evaluation of preoperative CT scan parameters may help identify patients who require access route preparation, as well as the best strategy to achieve it.

Paolo Alberto Del Sole, MD^1,2; Mattia Lunardi, MD, MSc³; Stefano Andreaggi, MD³; Simone Fezzi, MD, MSc³; Gabriele Pesarini, MD, PhD³; Roberto Scarsini, MD, PhD³; Flavio Ribichini, MD³

From the¹Department of Cardiology, SAOLTA Healthcare Group, Galway University Hospital, Health Service Executive and National University of Ireland Galway, Galway, Ireland; ²The Lambe Institute for Translational Medicine and CURAM, National University of Ireland Galway, Galway, Ireland; Division of Cardiology; ³Division of Cardiology of the Department of Medicine, University of Verona, Verona, Italy.

Dr Del Sole and Dr Lunardi contributed equally to the manuscript.

Disclosures: The authors report no financial relationships or conflicts of interest regarding the content herein.

Data availability statement: The data presented in this study are available on request from the corresponding author.  The data are not publicly available due to local policy regulating patients’ data sharing.

Address for correspondence: Mattia Lunardi, MD, MSc, University of Verona, Piazzale Aristide Stefani n°1, Verona 37126, Italy. Email: dott.lunardim@gmail.com; X: @Mattialunardi