Management of Lost Atherectomy Devices in the Coronary Arteries

ABSTRACT

Rotational and orbital atherectomy are important tools to treat calcific coronary disease. Entrapment of an atherectomy device, that is, rotational atherectomy burr or orbital atherectomy crown, is a serious complication during atherectomy. Loss of an atherectomy device is a more challenging complication that usually follows device entrapment. Although uncommon, given the increased adoption of atherectomy, it is important for interventional cardiologist to understand the mechanisms and principles for prevention and management of lost atherectomy devices. Given the paucity of reported cases, there is no clear consensus or defined approach as to how this complication is best managed. The current recommendations for burr/crown entrapment are generally to remove everything from the coronary vasculature and range from different percutaneous strategies and ultimately cardiac surgery. Conservative treatment with leaving the burr within the coronary vasculature has not been recommended. This article serves to review and provide a contemporary update on the management of lost atherectomy devices in the coronary arteries. We review the management and techniques centered around cases of lost atherectomy devices including a unique cohort of patients with lost burrs/crowns that were left in the coronary arteries. Finally, we provide an algorithmic approach to the contemporary management of lost atherectomy devices, incorporating a conservative strategy arm, and discuss situations where this may be considered.

1. Introduction

Despite significant advances in percutaneous coronary intervention (PCI), severe coronary calcification remains a challenge for interventional cardiologists. Up to 20% of patients undergoing PCI have moderate to severe coronary calcification [1]. Coronary calcification makes delivering wires, balloons, and stents difficult, limits lesion preparation with resultant stent underexpansion, and may reduce the efficacy of drug‐eluting stents (DES) [2]. As such, moderate to severe calcification has been associated with a higher rate of major adverse cardiovascular events (MACE) following PCI [3]. Tools for calcium modification before stent implantation include non‐atherectomy devices (scoring, cutting, super high‐pressure non‐compliant balloons, and intravascular lithotripsy) and atherectomy devices (rotational atherectomy [RA], orbital atherectomy [OA], and excimer laser coronary angioplasty [ELCA]). RA is a well‐established method for calcium modification and debulking. It utilizes a diamond‐encrusted elliptical burr that causes “differential cutting” and preferentially ablates hard, inelastic, calcified plaque that is unable to stretch away from the RA burr compared with healthy arterial tissue [4]. The newer OA system employs a 1.25 mm diamond‐coated eccentrically mounted crown which orbits in an elliptical path ablating tissue by exerting a centrifugal force on the vessel wall [5].

With a better understanding and development of modern atherectomy techniques, both RA and OA are increasingly used with low complication rates. Entrapment of the burr/crown within the calcified lesion is an uncommon, but serious complication during atherectomy [6]. The limitation with RA is that the proximal section of the burr does not have a diamond coating, and hence the atherectomy only occurs in an antegrade direction. Once it has slipped through the lesion, retrograde atherectomy is not possible when retracted. OA has the advantage of having a diamond coating on both the proximal and distal aspects of the crown, allowing bidirectional atherectomy if needed. With technical advances, PCI is being increasingly performed in complex calcific and tortuous coronary anatomy with a proportional increased risk of burr entrapment. An important complication following burr/crown entrapment is breakage and disconnection with subsequent loss of the atherectomy device in the coronary vasculature. Treatment algorithms for burr entrapment are described in the literature, usually with an intact driveshaft [7]. Complete loss of the atherectomy device in the coronary artery is uncommon, and to our knowledge, there are only three reported cases of complete burr loss including fracture of the Rotawire that were successfully retrieved percutaneously [8, 9, 10, 11] and two reported cases that required surgical removal [12]. Given the paucity of reported cases, there is no clear consensus or algorithmic approach on how this potentially catastrophic complication should be managed. Importantly, the principles of managing complete RA burr loss are naturally applicable to OA. To our knowledge, the loss of an OA crown has not been previously described.

The current recommendations for burr/crown entrapment are generally to remove everything from the vasculature and encompass different options including the use of guide extensions, trapping balloons, snares, subintimal techniques, and finally cardiac surgery. Conservative treatment with leaving the burr behind is not recommended [6, 13, 14]. Although an uncommon complication, given the increased adoption of atherectomy devices, it is important for interventional cardiologists to understand the principles for prevention and management. This article serves to provide a contemporary review and update on the management of lost atherectomy devices in the coronary arteries. We review the different management options and techniques centered around seven cases of lost atherectomy devices including a unique cohort of patients with lost burrs/crowns that were managed conservatively, that is, lost devices not retrieved and left safely within the coronary arteries. We provide an algorithmic approach to the contemporary management of lost atherectomy devices, incorporating a conservative strategy arm, and discuss situations where this strategy might be considered a safe alternative to open heart surgery.

2. Clinical Vignettes

2.1. Case 1—Lost Rota Burr and Retrieval (Figure 1)

Figure 1.

Lost rota burr and retrieval. (A) Tortuous and calcified proximal LAD. (B) broken RA 1.25 mm burr with short driveshaft remnant (C) rotational atherectomy driveshaft is constructed from 3 spirals/springs. (D) balloon dilatation to modify the calcium and dislodge the burr. (E) attempt to catch the burr with a wire wrap technique. (F) A guide extension was advanced over the wrapped wires, distal driveshaft and the burr. (G) successful burr removal by pulling back the wrapped wires together with the guide extension. (H) the burr and part of the distal driveshaft still attached to the burr entangled in the wrapped wires. (I) final angiographic result after stenting the proximal and mid LAD. [Color figure can be viewed at wileyonlinelibrary.com]

A 55‐year‐old lady had a background history of hypertension, hypercholesterolemia, type 2 diabetes, and peripheral vascular disease. She underwent emergency right femoral‐peroneal bypass and femoral endarterectomy procedure for acute right lower limb ischemia. She suffered a post‐operative non‐ST elevation myocardial infarction (NSTEMI) with dynamic anterolateral repolarization changes. Echocardiography documented severe impairment of LV systolic function. Given her comorbidities, she was declined for cardiac surgery and was referred back for percutaneous revascularization, PCI to the culprit left anterior descending coronary artery (LAD) was proposed. After crossing the diseased LAD with a Pilot 50 wire and exchanging it for a Rota Floppy wire, atherectomy was performed with a 1.25 mm burr (170,000 rpm, total ablation time 170 s). The proximal LAD was very tortuous and rigid owing to the calcium. There was burr entrapment, and the burr could not be retracted back from the LAD into the guide catheter. The burr itself was free and could be advanced antegradely. Applying gentle traction to retrieve the burr, resulted in breakage of the burr. The driveshaft had broken at the neck of the burr, leaving the burr tip with a 40 mm remnant of the Rota wire. After placing a parallel BMW Universal 2 wire, sequential balloon dilatations with a 2.0/15 and a 2.5/15 NC balloon were performed to dislodge the burr. This facilitated the placement of two additional BMW Universal 2 wires to proceed with a wire wrap technique [15, 16]. After placing the additional two wires, the three wires were twisted to wrap around the burr, driveshaft, and Rota wire remnant, a 5.5 F Guideliner was placed over the wrapped wires and delivered into the LAD artery just proximal to the burr. Applying traction, the wrapped wires, burr/driveshaft, and Rota wire remnant were drawn back into the guide extension and safely removed. Following this, the LAD was wired with a new BMW Universal 2 wire and following further predilatation, the LAD was treated with two DES. The patient has since undergone staged revascularization of the left circumflex artery (LCX), and the diffuse disease in the RCA has been managed conservatively. The patient has been asymptomatic and clinically stable at 1‐year follow‐up.

2.2. Case 2—Lost Rota Burr and Retrieval (Figure 2)

Figure 2.

Lost rota burr and retrieval. (A) RCA CTO. Dual injection setup. IABP and Swan‐Ganz catheter is seen. (B) RA burr detached completely at neck. Can see the driveshaft tip (white arrow) and detached burr on rota wire (black arrow). (C) 2.0/15 balloon over rota wire. Anchor to deliver 7 F guide extension. (D) Guide extension in distal RCA and brought to the lost burr with traction on intact rota wire. (E) Retract the whole system. (F) Final angiogram. Dissected vessel with flow. Left alone for conservative management. [Color figure can be viewed at wileyonlinelibrary.com]

A 58‐year‐old male with a background history of coronary artery bypass surgery (CABG) presented with an NSTEMI and decompensated heart failure. Coronary angiography documented an acute on chronic occluded RCA and the patient underwent PCI to RCA with intra‐aortic balloon pump support. After crossing the lesion with a Fielder XT‐A wire, the wire was exchanged for a Rota Floppy wire. RA was performed with a 1.5 mm burr at 150,000 rpm. During the final ablation run, the burr got stuck within the distal RCA, and became completely detached from the driveshaft. The driveshaft had broken at the neck of the burr, leaving the burr free in the distal RCA on an intact Rota wire. A 2.0/15 SC balloon was carefully passed over the Rota wire just proximal to the lost burr. This provided an anchor for the smooth passage of a 7 F Guideliner over the wire and delivered to the burr. The intact Rota wire prevented distal migration of the burr, and whilst applying traction on the Rota wire, the burr was wedged into the Guideliner. The Guideliner and Rota wire were both retracted as a single unit allowing successful retrieval of the lost burr.

2.3. Case 3—Lost Orbital Crown and Retrieval (Figure 3)

Figure 3.

Lost orbital crown and retrieval. (A) calcified lesion of the mid RCA and a subtotal occlusion of the distal RCA. (B) Successful antegrade wiring of the RCA. (C) Glide assisting OA crown beyond the proximal lesion. (D) Separation (white arrow) of tip and crown from the proximal drive shaft. (E) Unsuccessful attempt to parallel wire with Whisper LS using Turnpike LP (white arrow). (F) The driveshaft was removed, 6 F guide extension was advanced and with gentle tension, the guide extension (black arrow), Viperwire, and crown (white arrow) all came out together. (G) Final result after IVL and DES. (H) Final OCT. (I) Tip of the OA just behind the crown. (J) OA drive shaft without the tip. [Color figure can be viewed at wileyonlinelibrary.com]

An 82‐year‐old gentleman with a history of hypertension, hypercholesterolemia, type‐2 diabetes, and obstructive sleep apnea presents with angina. Coronary angiography documents a chronically occluded RCA and LCX. He underwent PCI of the RCA. After crossing the lesion with a Whisper MS wire, a Turnpike LP microcatheter was advanced. Given the significant calcification, atherectomy was proposed. The Whisper wire was replaced with the Viperwire, and OA was performed. The crown was moved past the proximal lesion under GlideAssist and treated with retrograde atherectomy followed by antegrade atherectomy. Next, the crown was moved past the mid RCA lesion under GlideAssist and the lesion was treated with retrograde atherectomy. As the device was being withdrawn, fluoroscopic imaging demonstrated that the driveshaft was disconnected from the crown. The Viperwire was intact. The driveshaft remnant was completely removed. Delivery of a parallel wire was difficult. A 6 F Guideliner was delivered over the Viperwire to the crown. Applying gentle sustained traction on the wire, the Guideliner, crown, and Viperwire were removed as a unit. The RCA was then wired with a Wiggle wire. The disease was then treated with a 2.5/15 mm Scoreflex scoring balloon and then IVL with a 4.0/12 mm shockwave C2+ balloon with the delivery of all 120 pulses in the proximal and mid RCA. Using optical coherence tomography (OCT) guidance, the disease was treated with two overlapping DES. The patient was discharged the next day and has been stable at 8 weeks follow‐up.

2.4. Case 4—Lost Rota Burr and Driveshaft in a CTO of the LCX (Figure 4)

Figure 4.

Lost rota burr and driveshaft in a CTO of the LCX. (A) CTO of the distal RCX. (B) broken and entrapped 1.25 mm RA burr with 12 mm of the driveshaft in the CTO body. (C) artifacts in the cardiac MRI. (D) inferolateral ischemia detected by cardiac MRI. (E) attempt to snare the driveshaft with a gooseneck 2 mm snare. (F) ADR and subsequent stenting past the broken burr and driveshaft.

A 61‐year‐old gentleman with a history of hypertension, hypercholesterolemia, type‐2 diabetes, renal transplantation, and previous NSTEMI was treated with PCI to LAD with two DES. He presented with angina and exercise‐induced ventricular ectopy. Echocardiography documented preserved left ventricular function (LVEF 50%) with inferolateral wall hypokinesia. Coronary angiography documented a chronic total occlusion (CTO) of the proximal LCX. After successfully crossing the CTO with a Fielder XT‐A wire and then exchanging it for a Rota floppy wire, RA was performed with a 1.25 mm burr. The vessel was heavily calcified requiring prolonged and extensive ablation runs (180,000 rpm, total ablation time 212 s). This ultimately resulted in breakage of the shaft with loss of burr and 12 mm of the driveshaft left behind in the coronary artery. The Rota wire position was lost with the retrieval of the remaining shaft and the burr left embedded in the body of the CTO body. An attempt to snare the broken driveshaft was unsuccessful due to the inability to push the snare over the remaining distal driveshaft. During retrieval attempts the ostial LAD became dissected and required treatment with a single DES. Repeat attempts to retrieve the burr and driveshaft using the knuckle‐twister technique were unsuccessful. The burr and driveshaft remnant were left behind in the CTO body since none of the parts were in contact with any open vessel and the patient was discharged in a stable condition. However, his symptoms persisted despite optimal medical therapy. A follow‐up cardiac magnetic resonance imaging (MRI) was performed to assess viability and ischemia. This was performed safely, and although the driveshaft and burr remnant resulted in extensive artifacts, ischemia and viability could be demonstrated. Thus, he underwent a repeat attempt to recanalize the CTO. Repeat angiography documented the burr in the CTO body with no evidence of any migration. As the burr was heavily embedded in the calcific CTO body, an antegrade dissection reentry strategy was proposed to work around the calcific CTO body and driveshaft/burr remnant. The CTO body and driveshaft remnant were successfully bypassed with the extra‐plaque passage using a Conquest Pro 12 and Gladius EX wire, and finally a single DES was deployed in the proximal LCX with complete exclusion of the burr/driveshaft remnant behind the stent, leaving it embedded in the vessel wall. The patient has been asymptomatic and clinically stable at 4‐year follow‐up.

2.5. Case 5—Lost Rota Burr in a Tortuous RCA (Figure 5)

Figure 5.

Lost rota burr in a tortuous RCA. (A) Tortuous and heavily calcified RCA. Limited access due to Sapien XT valve. (B) Successful wiring with a polymeric guidewire, exchange for a rot floppy wire with the use of a microcatheter. (C) RA with a 1.25 mm burr at 180,000 rpm Breakage of the burr from the driveshaft at the third and acute bend. (D) Successful retrieval of the burr up to the proximal RCA with multiple balloon inflations and pullbacks. (E) Multiple failed attempts to retrieve the burr with the knuckle twister technique. Parts of the Fielder XT‐A were lost in the RCA. (F) Stenting of the proximal RCA jailing the broken guidewire and the burr in the wall. Followed by multiple postdilatations. (G) Final result with stenting of the entire RCA burying the lost equipment in the vessel wall. (H) Angiographic follow‐up after 3 months. (I) OCT at follow‐up shows good stent expansion, apposition, and healing with a sufficient MSA. The burr is deep in the vessel wall undetectable by OCT. [Color figure can be viewed at wileyonlinelibrary.com]

A 77‐year‐old gentleman with a background history of hypertension, hypercholesterolemia, type‐2 diabetes, and previous transcatheter aortic valve replacement (TAVR) for severe aortic stenosis presented with angina. Coronary angiography demonstrated severe calcific disease in a large dominant RCA and PCI to the RCA was proposed. After the initial tortuous passage of a Fielder XT‐A wire, the wire was exchanged for a Rota Floppy wire. RA was then performed with a 1.75 mm burr. The vessel was heavily calcified requiring prolonged and extensive ablation runs (180,000 rpm, total ablation time 235 s). During the ablation, there was a breakage of the burr from the driveshaft in the mid RCA. The driveshaft had broken at the neck of the burr. In the process of retrieving the driveshaft, the Rota wire position was lost, with a completely free burr in the mid RCA. Attempts to retrieve the burr with gooseneck snares were futile as no driveshaft/wire remnant would otherwise facilitate snaring, and the snare was not able to capture the isolated burr. The burr was partially retrieved up till the proximal RCA with a slow and controlled retraction of a distally inflated balloon. Finally, attempts were made to retrieve the burr from the proximal RCA using the knuckle‐twister technique with multiple Fielder XT‐A wires. Still, this was unsuccessful due to the absence of the driveshaft and Rota wire remnant. In this process, a Fielder XT‐A tip was broken and lost in the proximal RCA. Further attempts to retrieve the burr were abandoned as with these aggressive attempts to retrieve the burr, there was an Ellis type 1 perforation in the proximal RCA. After multiple balloon dilatations, the RCA was treated with three DES with complete exclusion of the burr and fractured wire remnant behind the stent, leaving them embedded in the vessel wall. The perforation had sealed and the patient was discharged in stable condition. Follow‐up coronary angiography at 3 months demonstrated the burr to be embedded in the vessel wall without any migration. An OCT study documented a well‐expanded stent with good endothelialization and was unable to detect the buried burr indicating a deep position within the vessel wall. The patient has been asymptomatic and clinically stable at 2‐year follow‐up.

2.6. Case 6—Lost Orbital Crown in Diagonal Branch (Figure 6)

Figure 6.

Lost orbital crown in diagonal branch. (A) tortuous and calcified large diagonal branch. (B) stuck OA crown after the first bend of the diagonal branch. (C) forced pushing, pullback. (D) driveshaft breakage with the driveshaft/crown and Viperwire remnant in the diagonal branch. (E) occluded side branch by the driveshaft/crown remnant.

An 86‐year‐old lady with a background history of hypercholesterolemia, cerebrovascular disease, and previous PCI to the RCA and PCI of an LAD CTO with three DES was troubled with escalating anginal symptoms. She had residual severe disease in a large diagonal branch (D1) and given her persisting anginal symptoms, PCI to the D1 was proposed. The D1 was a calcified tortuous vessel. After delivering a Viperwire, OA was performed at 80,000 rpm. Advancement of the OA crown was challenging owing to the tortuosity, requiring aggressive forward push. However, this resulted in progressive retraction of the wire. Eventually, the wire had come back far enough for the distal spring tip of the wire to interact with the driveshaft tip, ultimately causing the device to stall with subsequent driveshaft and wire fracture. The wire and driveshaft containing the crown were completely contained within the D1. There was no remaining flow in the D1 due to the lost device. As the patient was asymptomatic, and no device or wire fragments extended back to the LAD, no further action was taken. The lost device and wire fragment were left in the D1, and the procedure was stopped. The patient remained asymptomatic and clinically stable during a follow‐up visit 1.5 years later.

2.7. Case 7—Lost Rota Burr in Distal RCA (Figure 7)

Figure 7.

Lost rota burr in distal RCA. (A) Undilatable lesion. (B) Ostial RA. (C) Driveshaft breakage, burr (black arrow), driveshaft (white arrow). (D) lost RA burr. (E) Dilatation of lesion. (F) Distal balloon inflation and retrieval. (G) Unsuccessful retrieval of the burr (white arrow) into a guide extension (black arrow). (H) Distal embolization of the burr (white arrow) into the PDA.

An 82‐year‐old female with a background history of previous CABG surgery presented with symptomatic severe aortic stenosis and good left ventricular function. She has severe calcific RCA disease that has not previously been grafted. She was turned down for redo surgery and was referred for PCI of the RCA before TAVR. Given the severe calcification in the proximal RCA, PCI with RA was proposed. After crossing the lesion with a BMW wire, this was exchanged for a Rota floppy wire. RA was performed with a 1.75 mm burr (170,000 rpm, total ablation time 170 s). After successful passage through the calcific lesion, there was burr entrapment after the first bend in the RCA. After parallel wiring with a BMW wire and sequential balloon inflations in the proximal RCA, the trapped burr could not be released. Eventually, the shaft of the burr was pulled to retrieve the burr resulting in breakage of the burr. The driveshaft had broken at the neck of the burr, leaving the burr in the proximal RCA just distal to the lesion with an intact Rota wire. It was unclear whether there was any driveshaft left with the burr, and thus snaring was attempted to grasp any potential driveshaft remnant, but this was not successful. Acknowledging that the burr could not be retrieved from the proximal RCA, the burr was then pushed down into the distal posterior descending artery (PDA) by using an inflated balloon. The RCA was extensively dissected in its proximal and mid‐section, but given that there was TIMI 3 flow, it was managed conservatively. The patient had a good recovery from the procedure and was discharged in stable condition. She subsequently underwent TAVR, and the RCA disease has been managed conservatively. She has been asymptomatic and clinically stable at 2‐year follow‐up.

2.8. Case 8—Lost Rota Burr at a Bifurcation (Figure 8)

Figure 8.

Lost burr at bifurcation. (A) Tortuous calcified medial RCA. (B) Broken driveshaft at the neck of the burr (white arrow). (C) Detached 1.5 mm burr and retracted driveshaft (white arrow). (D) Ping‐pong and mobilization attempt using a 2.5/15 NC balloon. (E) Failure to snare (white arrow) the burr. (F) Distal embolization and occlusion of the PDA (dashed line). (G) Exclusion from the circulatory by a stent. (H) Free flow in all distal branches, including the PDA. (I) Fully expanded stent pushing the burr into the vessel wall.

An 85‐year‐old female presented to the emergency department with persistent chest pain and was diagnosed with an NSTEMI. Coronary angiography demonstrated three‐vessel CAD, and primary PCI of the culprit lesion in the proximal RCA was undertaken. Following successful wiring with a Rotawire Drive, RA was initiated using a 1.5 mm burr. During advancement, the burr became lodged at the first curvature of the proximal RCA. Fluoroscopy revealed a gap in radiopacity, indicating driveshaft fracture at the burr neck. Attempts to retrieve the device resulted in the burr detaching and remaining within the artery. Concurrently, the Rotawire was found to be fractured, precluding burr retrieval. A parallel wire was engaged, and attempts were made to mobilize the burr using a 2.5/15 NC balloon. However, retrieval with a Goose Neck snare was unsuccessful. Subsequently, the burr embolized distally to the crux cordis, occluding the PDA and causing recurrent chest pain with signs of inferior ischemia. A workhorse wire was successfully advanced into the posterolateral branch, and a stent was deployed across the bifurcation, effectively jailing the burr and restoring antegrade flow. Ischemia and chest pain promptly resolved. ClearStent imaging confirmed optimal stent expansion with complete entrapment of the burr. The proximal RCA was treated with a drug‐eluting balloon. The patient remained asymptomatic following the procedure but succumbed to severe pneumonia 5 days later.

3. Discussion

The above cases demonstrate different clinical scenarios with lost atherectomy devices including RA burrs and OA crowns. While lost RA burrs have previously been reported [8, 9, 10], we have reported the first cases of lost OA crowns. With OA, the ability to perform retrograde ablation is considered an important advantage over RA, theoretically reducing the risk of crown entrapment and thereby device loss. However, the presented cases highlight that the risk of entrapment and device loss is not completely mitigated with the newer OA system. Cases 1−3 demonstrate the different percutaneous strategies that can be used to retrieve the lost device. Case 1 is a unique case with the first report of successful retrieval of a lost burr using a wire wrapping technique. However, when the device cannot be removed, a conservative approach may be considered. Cases 4−8 demonstrate unique cases where retrieval was not possible, and a conservative approach was adopted where the atherectomy devices were left in situ within the coronary artery. Cases 3 and 6 provide the first‐ever reports of lost OA crowns and demonstrate both percutaneous removal and conservative strategies. Table 1 provides a summary of all the cases. To our knowledge, the current article provides the first description of conservative management for lost devices and demonstrates the feasibility and safety of such an approach with both types of atherectomy devices. The strategies for percutaneous retrieval of the lost device depend on the site of the driveshaft breakage and the integrity of the underlying atherectomy wire. Thus, it is important to have a basic understanding of the structure, components, and make‐up of the atherectomy devices to help understand the mechanism of device loss and guide different management strategies. When percutaneous retrieval fails, conservative management can be an option in certain situations. Surgery should be considered when driveshaft or guidewire remnants cannot be retrieved and remain in the aorta or cause coronary vessel obstruction or ischemia.

Table 1.

Summary of all the cases, type of atherectomy used, mode of failure and type of treatment strategy.

Case	Lost device	Wire intact	Driveshaft remnant	Retrieval possible	Successful retrieval method	Conservative approach	Conservative strategy
1	RA burr	No	Yes	Yes	Wire wrap technique	No	—
2	RA burr	Yes	No	Yes	Wire traction with GE	No	—
3	OA crown	Yes	No	Yes	Wire traction with GE	No	—
4	RA burr	No	Yes	No	—	Yes	Stenting the burr/driveshaft into the vessel wall.
5	RA burr	No	No	No	—	Yes	Stenting the burr into the vessel wall.
6	OA crown	No	Yes	No	—	Yes	Crown/driveshaft left within the occluded vessel.
7	RA burr	No	No	No	—	Yes	Balloon‐facilitated migration to distal branch.
8	RA burr	No	No	No	—	Yes	Stenting the burr into the vessel wall.

Abbreviations: OA, orbital atherectomy; RA, rotational atherectomy.

3.1. Structure of Atherectomy Devices

It is important to have a basic technical understanding of atherectomy devices and the dedicated wires upon which they are delivered to help guide different percutaneous strategies for retrieval in cases of burr/crown loss and to determine when it might be appropriate to leave these devices for conservative management. Schematic diagrams of the structure of the atherectomy devices are shown in Figure 9.

Figure 9.

Structure of atherectomy devices. (A) Rotational atherectomy device. Inbox: most frequently used burr sizes in mm. (B) Orbital atherectomy device. One single crown of 1.25 mm is available for coronary application. [Color figure can be viewed at wileyonlinelibrary.com]

3.1.1. RA

There are three components to using RA. The first component is the use of a dedicated 0.009″ 325 cm stainless‐steel Rota wire with a 0.014″ radiopaque platinum flexible coiled tip. The wire tip is a safety feature preventing the burr from advancing beyond the wire. The second component is the nickel‐plated brass elliptical burr with diameters varying from 1.25 to 2.5 mm. The distal end of the burr is encrusted with diamond particulates. The proximal end of the burr is not encrusted with diamond particulates and is welded onto the drive shaft, which slides over the Rota wire. The driveshaft is comprised of three coiled stainless‐steel wires. When breakage occurs, it can happen anywhere along the drive shaft, but the weakest point is the neck, where it is laser welded onto the burr. As such the common sites for breakage are at the neck or in the immediate proximal driveshaft segment. Traction of the driveshaft can also result in stretching before physical breakage occurs. A Teflon sheath over the drive shaft minimizes injury to the artery. This also allows cooling and lubrication during the ablations with the infusion of different solutions. The third component is the advancer unit connected to an air turbine. Within the console is a high‐pressure air turbine connected to the drive shaft [1].

3.1.2. OA

There are three components to using OA. The first component is a dedicated 0.012″ 335 cm nitinol Viperwire with a 0.014″ radiopaque distal spring tip. Similar to the Rota wire, this tip is a safety feature preventing the crown from advancing beyond the wire. The second component is an eccentrically mounted crown 1.25 mm in diameter, and placed 6.5 mm behind the tip of the drive shaft. An abrasive coating consisting of diamond particulates is adhered to the tungsten crown base with a nickel alloy. The exposed diamond particulates provide the sanding treatment while the crown rotates and orbits. The driveshaft is comprised of three coiled stainless‐steel wires. The crown base material is either stainless steel or tungsten alloy and can be operated at two speeds 80,000 and 120,000 rpm. The crown is generally secured to the distal end of the driveshaft using a sleeve‐and‐key or collet system, which provides a rigid, yet secure, connection. Other than for RA, the common site of breakage in OA is not at the neck but in the immediate proximal driveshaft segment due to the missing welding point. Traction of the driveshaft can also result in stretching before physical breakage occurs. Similar to RA, a Teflon sheath over the drive shaft minimizes injury to the artery and allows for cooling and lubrication during the ablations with the infusion of Viper Slide Solution. The third component is the advancer unit, which is electrically powered and rotates the driveshaft at two different speeds, which in turn rotates the eccentric crown to generate an orbital motion [17].

3.2. Mechanism of Burr/Crown Loss

Burr/crown loss is defined as a free burr/crown in the coronary circulation that is completely detached from the driveshaft of the atherectomy device. The mechanism of burr/crown loss with atherectomy devices can be better understood when dissecting the process into two stages: (1) burr/crown entrapment; and this is followed by (2) burr/crown detachment from the driveshaft.

3.2.1. Burr/Crown Entrapment

In general, the mechanisms for device entrapment are similar for both RA and OA. However, some factors are specific to RA. Device entrapment is more common for RA burrs, as the ablative diamond coating at its distal surface for antegrade ablation, but the proximal part is smooth without diamonds, which prohibits backward ablation. Burr advancement beyond the calcific lesion without significant debulking is often seen with the smaller 1.25 mm burr. A small burr can be advanced beyond a heavily calcified plaque before sufficient ablation, especially when the burr is pushed firmly at high rotational speed. During high‐speed rotation, the frictional heat may cause vasodilatation enlarging the space between plaque and burr, and with the lower coefficient of friction during motion, the burr may pass the calcified lesion easily without debulking a significant amount of calcified tissue. In this situation, the ledge of calcium proximal to the burr may prevent burr withdrawal, and often the small burr remains free beyond this in the distal vessel. The 1.25 mm burr is unique in design compared to all other burr sizes in the way that it features a larger non‐diamond coated contact surface on its side, which makes the small burr more vulnerable to this complication. RA runs at lower speeds (< 140,000 rpm) can predispose to greater decelerations and burr stalls within the calcific lesion. Factors that are common for both RA and OA include: (a) advancement of the burr/crown before sufficient ablation resulting in the device becoming completely stuck within the calcific lesion. Forceful aggressive burr advancements result in greater decelerations, and decelerations > 5000 rpm are associated with burr stalls and entrapment. However, if the burr/crown can be advanced beyond the disease and remains free, then OA may be advantageous in this scenario with its backward ablation; and (b) anatomical factors, including vessel tortuosity, angulated vessels, eccentric lesions, and long lesions. These increase the coefficient of friction during the ablation, which may predispose to entrapment.

3.2.2. Burr/Crown Detachment

Following burr/crown entrapment, burr/crown loss occurs following complete detachment from the driveshaft. The driveshaft is not composed of a solid material but is composed of three coiled stainless‐steel wires which give the driveshaft its flexibility while being able to provide high‐speed rotation. Breakage of the driveshaft can occur in two ways. First, is due to excessive torsion of the driveshaft. During the process of entrapment, the driveshaft is initially rotating at the designated high ablation speed, but as the burr/crown rotation comes to a rapid halt, the relative increased rotation of the driveshaft with respect to the burr/crown for a short duration before a complete stall, results in an exaggerated twisting process with high torque. This leads to excessive torsion and ultimately breakage of the driveshaft either at the neck where it is attached to the burr or within the drive shaft itself just proximal to the burr/crown. Additional stress can be exerted by manual traction of the driveshaft. Manual traction is often used as a measure to remove the trapped burr/crown, with variable success. When this is not successful, the tension in the driveshaft, causes it to initially elongate and ultimately to break. In cases of RA when the burr may still be free distal to the calcific disease, occasionally traction may be applied whilst applying Dynaglide rotation, with the intention that Dynaglide rotations will reduce the coefficient of friction that may allow the burr to be retrieved with traction. However, if the traction is applied while applying Dynaglide rotation and the burr is not successfully retrieved, the burr becomes stuck and comes to a halt in the calcific lesion. In this situation, the excessive torsion of the driveshaft will further facilitate breakage of the driveshaft.

3.3. Retrieval of Lost Burr/Crown

When the burr/crown is lost in the coronary circulation, retrieval is recommended. An algorithm for managing a lost burr/crown is shown in Figure 10. There are two important steps to help determine the strategies for removal. First and foremost, the broken driveshaft and atherectomy device that is retrieved must be carefully examined to determine the site of breakage and to determine whether any length of the driveshaft is still attached to the burr/crown. Secondly, it is important to determine if the Rota or Viperwire is intact. Often these wires break proximal to the burr/crown, but in some cases, remain completely intact. If the wire is broken, then it is important to determine the length of the remaining wire fragment and whether the wire remnant and its tip contain the lost burr/crown. The latter point is important as the larger 0.014″ tip of the dedicated Rota and Viperwires trap the burr/crown on the wire and ensures that it does not travel more distally. The presence of the wire through the burr/crown or a driveshaft remnant attached to the burr/crown is important, as this provides possibilities for securely capturing the lost crown/burr using different strategies [11]. In general terms, if there is no driveshaft or wire remnant, then retrieval becomes challenging, and in those circumstances, a conservative approach may be considered. To be able to successfully retrieve the device, there are two important prerequisites. First, the burr should be retractable and should be freed from the underlying calcific disease. Secondly, the device should be capturable, and the driveshaft or wire remnants facilitate this. If the impacted device is not free, then parallel wiring followed by parallel balloon inflations can dislodge the device from the calcific disease. If this fails, or if parallel wiring is challenging, then subintimal techniques can be employed. In this scenario, a polymer jacketed wire can be used to enter extra‐plaque space to facilitate delivery of balloons and allow external crush of the calcific disease, in the hope of dislodging the device. Rarely, laser atherectomy may be introduced over a parallel wire to modify and soften the underlying plaque in the hope of subsequently dislodging the device from the underlying diseased vessel. The next step is the successful capture of the device. If the wire remains completely intact, given that the distal tip is larger and prevents distal migration of the device, just pulling the wire at this stage may allow the device to be retrieved. If this simple maneuver does not work, the passage of a guide extension over the wire may allow greater controlled traction and retrieval of the device. Before introducing a guide extension over the RA or OA device, the driveshaft has to be cut off close to the advancer unit, and the Teflon sheath must be removed. If this fails, the traction force can be increased by inflating a small balloon within the guide extension system to trap the wire and then retracting the guide extension, wire, and device as a single unit. If this does not work, then aggressive traction of the wire should be avoided as this may break the otherwise intact wire. In this scenario, to further increase the traction, the knuckle‐twister or wire wrapping techniques can be applied. The knuckle‐twister technique is particularly useful to capture the device when there is a small wire remnant through the device or with a driveshaft. This technique requires delivering a Fielder XT‐A guidewire with a backflipped tip (knuckle) alongside the lost device and its driveshaft/wire remnant. The guidewire is purposely rotated and the open loop closes and tightens around the lost device and driveshaft/wire remnant [18]. Delivering a knuckled wire alongside the lost device may be challenging. In this situation, a parallel non‐knuckled wire, which may be easier to deliver, can be placed alongside the device. A knuckled Fielder XT‐A wire can then be delivered over the parallel wire using a dual‐lumen microcatheter system. If the delivery of knuckled wires is difficult, an alternative solution is to deliver two non‐knuckled parallel wires alongside the lost device. Wire‐wrapping is then induced by purposely rotating the wires [15, 16]. Whichever technique is ultimately used, once the lost device and driveshaft wire remnant are securely twisted and wrapped, this can then be retracted and removed as a unit. The use of a guide extension is particularly useful and recommended in this setting. This provides a more controlled and stronger traction while at the same time isolating the entity and reducing friction in the proximal disease. An alternative to the knuckle‐twister or wire‐wrapping technique is to use microsnares to capture the driveshaft/wire remnants. Finally, if there is no driveshaft/wire remnant and the burr/crown is completely free in the coronary artery, then retrieval can be challenging. In some cases, a free crown/burr may be retrieved by careful retraction of a distally inflated balloon. Where the free crown/burr is in proximal vessels with relatively little upstream disease, a balloon may be inflated over a parallel wire distal to the free crown/burr and carefully withdrawn back to bring the device into the guiding catheter [8, 9]. For more distal locations, the same strategy may be employed but with the use of a guide extension delivered into the distal vessel proximal to the lost device. This strategy is more likely to be successful with the retrieval of burrs as opposed to crowns, as burrs have larger cross‐sectional profiles increasing the likelihood of capture with balloon retraction. The use of a distal embolic protection device to retrieve a lost burr has also been described [10]. Here, the embolic protection basket is deployed distal to the device and retracted carefully to capture and retrieve the lost device. However, given the underlying partially ablated calcific disease, this strategy should be used with caution, as the retrieval basket itself may become entrapped, further compounding the problem. Indeed, capturing the device without a driveshaft/wire remnant is very difficult in these scenarios, and a conservative management should be considered.

Figure 10.

Algorithm for percutaneous and conservative management. Algorithm for management of atherectomy device entrapment and loss. [Color figure can be viewed at wileyonlinelibrary.com]

3.4. Conservative Management

When the burr/crown cannot be retrieved, then conservative management appears to be the next best option. If the device is embedded within a diseased segment where there is no risk for further distal embolization, it can be left behind. If the burr/crown has the potential to migrate or embolize distally or if it is compromising antegrade flow with resultant ischemia, then the device can be fixated to the vessel wall by delivering a parallel wire followed by ballooning and stenting. It is important that the burr/crown and any driveshaft remnant attached to it are covered and completely fixated to the vessel wall. If a parallel wire cannot be delivered easily within the intraplaque space, then subintimal techniques can be utilized, to work around the lost device and to exclude it from the coronary circulation. The main concern about leaving a device within the vessel wall is the potential for it to compromise the stent and/or the potential for migration into deeper cardiac structures and/or the pericardium. The cases above have demonstrated the safety of the conservative approach with clinical stability at long‐term follow‐up. The angiographic follow‐up in the above cases has shown that following conservative management, the device does not appear to migrate with no compromise in stent patency. The angiographic follow‐up for the cases ranged from 6 to 24 months. The OCT follow‐up in one patient after 6 months was reassuring, with good endothelialization of the stent and vessel healing. It is also important to note that MRI was also feasible, and although cardiac MRI may be limited by extensive artifacts, ischemia testing was still possible. Although subsequent migration of the device beyond 12 months cannot be excluded, it is certainly reassuring to note clinical stability at longer‐term clinical follow‐up. In cases where angiographic follow‐up is not possible, given the paucity of reported cases and data in this field, an interval follow‐up CT scan may be useful to ensure the device has not migrated.

3.5. Surgical Management

Surgery has been recommended to remove the lost burr/crown when it cannot be retrieved percutaneously. With the diverse percutaneous strategies for removal and the demonstrated safety of leaving the device within the coronary artery, the role of surgery appears minimal. However, surgery may still be considered where the burr removal is deemed to be necessary for the rare scenario where it may be causing ischemia, and the vessel cannot be treated with a stent. Any wire or driveshaft remnant extending to the aorta and the inability to remove them percutaneously will also require surgery.

4. Conclusions

Loss of an atherectomy device is a rare but serious complication of atherectomy, that usually follows device entrapment. The device can often be removed percutaneously using different strategies based on the site of device breakage and the absence/presence of the guidewire through the lost device. However, if the device cannot be retrieved and there are no concerns about ischemia or distal embolization, then conservative management is the next best option where it can either be left in the coronary circulation or contained and fixed to the vessel wall. A conservative approach appears to be safe with no adverse clinical outcomes at angiographic and clinical follow‐up. Surgery should be considered as a last resort in the rare scenario, where the lost atherectomy device cannot be retrieved percutaneously and poses a risk for ischemia (Central illustration 1).

Central Illustration 1.

Lost atherectomy devices are rare but serious complications of coronary atherectomy. This review outlines current management strategies, including percutaneous and surgical approaches, and introduces a novel conservative strategy. An algorithmic approach is proposed to guide interventional cardiologists in the prevention and management of lost burrs or crowns during coronary interventions.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Appendix.

Illustration of conservative treatment scenarios

1.

A. jailing the lost device including the driveshaft behind a Stent. B. jailing the lost device without the driveshaft behind a Stent. C. leaving the lost device within the CTO body. D. pushing the lost device to a very distal vessel.