PART III - DIRECTIONAL ATHERECTOMY
TABLE OF CONTENTS
BOOKSHOP
Updated on November 20, 2020
PART III

Directional atherectomy

Antonio Colombo, Azeem Latib

Summary

The concept of atherectomy, defined as excision and removal of atherosclerotic plaque by a transcatheter technique, was introduced by Simpson et al in 1985. Thereafter, the technique underwent several generational improvements. Intuitively, directional coronary atherectomy prior to stenting should be complementary in reducing restenosis since in-stent restenosis is mainly due to neointima formation, which is increased proportionally to the amount of plaque. Furthermore, plaque removal prior to stent implantation may allow greater and more uniform stent expansion, and minimise plaque shift, thereby improving angiographic and clinical results. Several single-centre and multicentre registries supported the concept that directional atherectomy prior to stenting would reduce restenosis. However, 2 randomised trials (AMIGO and DESIRE) failed to substantiate the findings of these registries. The introduction of drug-eluting stents has further decreased the utilisation of atherectomy. Nevertheless, recent device registries have suggested that the combination of drug-eluting stents and directional atherectomy may reduce restenosis in complex lesions such as left main bifurcations. Careful lesion selection and optimal atherectomy are essential in preventing complications and determining the efficacy of directional atherectomy in improving angiographic outcomes. There may still be a role in the drug-eluting stent era for performing lesion debulking with atherectomy prior to stent implantation in complex lesions such as aorto/ostial lesions, bifurcations and left main lesions which have a large plaque burden.

Introduction

John Simpson first introduced the concept of removing obstructive tissue by a catheter-based excision technique [1, 2]. The idea was to overcome several limitations of conventional balloon angioplasty (PTCA). The first directional atherectomy of human vessels was performed in a superficial femoral artery in 1985 and in human coronary arteries the following year [1, 2, 3]. The peripheral device was approved by the Food and Drug Administration (FDA) in 1987 and the coronary device in 1990 [1, 2, 3, 4, 5]. By removing obstructive plaque, it was hoped that this debulking would reduce the high rates of restenosis observed with balloon angioplasty. Despite the initial enthusiasm for this first non-balloon percutaneous coronary interventional device, the early randomised clinical trials of directional coronary atherectomy (DCA) failed to demonstrate a clinically significant reduction in late angiographic or clinical restenosis after DCA versus PTCA [6, 7, 8, 9, 10]. These negative findings for DCA, combined with the approval of the Palmaz-Schatz™ coronary stent (Cordis Corporation, Miami Lakes, FL, USA) in 1994, resulted in a rapid decrease in the usage of this unique technology.

The most promising results with DCA were obtained when combining DCA with stenting [11, 12, 13, 14, 15]. Intuitively, DCA and stent implantation should be complementary in reducing restenosis as they act on different mechanisms of restenosis. In-stent restenosis is mainly due to neointima formation [16, 17], while late lumen loss following DCA is mainly due to negative arterial remodelling and to a lesser extent cellular proliferation [18, 19, 20, 21, 22, 23]. Therefore, at least theoretically, there is a pathophysiological basis for a combined strategy of DCA with stenting. Intravascular ultrasound (IVUS) studies have shown that both pre-intervention and post-intervention plaque burden are predictors of restenosis after stent implantation and most neointimal hyperplasia occurs at the site of the underlying plaque [24, 25, 26, 27, 28]. Plaque burden may also play an important role in the restenosis process by amplifying negative remodelling [9]. Furthermore, the presence of a large plaque burden may prevent optimal stent expansion despite the use of high inflation pressures or large balloons. This resultant stent underexpansion has been shown to predispose to restenosis even with drug-eluting stents (DES) [21, 29]. Debulking may also minimise plaque redistribution during coronary stenting (“snowplough effect” or ‘’plaque shift’’) and thus decrease PCI-related side branch occlusions and infarctions [30]. As stents reduce restenosis by opposing late negative remodelling, it was reasonable to assume that optimal lesion debulking (to minimise vessel wall trauma during optimal stenting) followed by stent implantation would be a promising approach to improve long-term clinical outcomes [9, 21, 23].

DCA equipment and technique

The prototype of the DCA catheter was the Simpson Coronary AtheroCath™ (Devices for Vascular Intervention Inc., Redwood City, CA, USA), which consisted of a metal housing with a fixed support balloon, a nose cone collection chamber, and a hollow torque tube, which accommodated a 0.014-inch guidewire. A cup-shaped cutter inside the housing was attached to a flexible drive shaft and was activated by a hand-held battery-operated motor drive unit. A lever on the motor drive unit allowed the operator to advance the cutter slowly through the lesion as it rotated at 2,000 revolutions per minute (rpm). Excised atheroma was stored in the distal nose cone collection chamber. The Simpson AtheroCath™ was the first device after balloon angioplasty to receive FDA approval for coronary intervention.

The second generation AtheroCath (SCA-EX™) was characterised by an improved nose cone design, and the third generation device (SCA-GTO™) had a redesigned shaft with better support and torque control than the EX. The Bantam™ catheter had a smaller shaft allowing 9 Fr guide catheter compatibility.

A more recent device is the Flexi-Cut™ (Guidant Corporation, Indianapolis, IN, USA). It has a cutter coated with titanium nitride ( Figure 1) and a shaft diameter slightly smaller than 6 Fr (0.076”, 1.94 mm). The shaft has not increased in size with larger devices. Rather, increasing diameter of the opposing balloon (2.5, 3.0, and 3.5 mm) adapted this device to larger vessels. As a result, the Flexi-Cut™ device was compatible with 8 Fr large lumen (>0.87” internal diameter) guide catheters. The rigid part of the cutting chamber was also shorter than in the GTO catheter (13 mm vs. 16 mm), improving tip flexibility while maintaining the length of the cutting window (9 mm).

The SilverHawk™ Plaque Excision System (FoxHollow Technologies, Inc., Redwood City, CA, USA; ev3 Inc., Plymouth, MN, USA) is a new atherectomy device which uses a different concept ( Figure 2). This device received CE mark approval for use in coronary arteries in December 2002 and is currently available in Europe but not in the USA where its indication is still limited to peripheral interventions. A family of three different SilverHawk™ catheters (compatible with both 6 Fr and 8 Fr guiding catheters) is available for coronary use. The SilverHawk™ catheter does not need an external balloon to establish vessel wall contact, and can be advanced over a short PTCA wire as a monorail device. When the motor drive unit is activated and the positioning lever is retracted, the distal portion of the cutter deflects, bringing the device into apposition with the lesion and exposing the inner rotating cutter. While the blade is spinning, the entire catheter is slowly advanced across the lesion, “shaving” the occlusive material from the artery and capturing the excised tissue in the tip of the catheter. Advancing the positioning lever, bringing the inner cutter back into the housing and restoring the “non-deflected” configuration completes the cutting process. Preliminary clinical experience with the SilverHawk™ catheter showed that plaque debulking can be performed safely and effectively in relatively small vessels and complex lesions located in mid-distal artery segments, with 6-month clinical outcome similar to prior atherectomy devices [31]. The SilverHawk™ technology, following the acquisition by ev3 Inc., has been further refined with the TurboHawk™ Super Cutter (ev3 Inc.) to deal with calcified plaque. It is used exclusively in peripheral arteries.

A new DCA device, the AtherCut (Nipro Co., Osaka, Japan) is being used in Japan. The device is similar in design structure to the Flexi-Cut but has some important improvements: nose cone was changed to a tapered type, outer diameter was minimized (1.95 mm, 7Fr guiding catheter compatible), the cutter was changed to a diamond-like carbon coating, the balloon surface on the vessel wall was changed to a flat type and the rotational speed was modified to 6,000 rotations per minute. These changes should improve device crossability, torque response and cutting performance [66].

Performing DCA correctly and safely requires formal training with direct supervision. Incorrect use of DCA can lead to major complications, which may not be resolved easily. Correct guide catheter choice is important as major dissections can occur due to incorrect manipulation. A supportive guidewire is essential for the Flexi-Cut™ device but is not necessary for the SilverHawk™ which can be used with any type of guidewire. In some situations DCA can be the only way to gain access to a side branch by removing obstructive plaque in the main branch. In any challenging anatomical setting, the evaluation of the artery size and results of atherectomy with IVUS should be considered. Even with contemporary DCA devices, calcium remains a major limitation and contraindication if extensive. The device should never be forced through a lesion: instead rotation should be employed. If there is unfavourable alignment, then the operator must accept that DCA cannot be performed. DCA should also be avoided in extremely angulated lesions due to the risk of coronary perforations.

FOCUS BOX 1DCA equipment and technique
  • Flexi-Cut™ and SilverHawk™ are the currently available and utilised devices
  • Flexi-Cut™ uses an external balloon to establish vessel contact while the SilverHawk™ uses a lever system to bring the device in contact with the lesion
  • Performing DCA correctly and safely requires formal training with direct supervision as incorrect usage can lead to major complications
  • Correct guiding catheter choice and manipulation as well as the usage of support guidewires are essential for safe and effective DCA
  • DCA should be avoided in heavily calcified or extremely angulated lesions

Performing optical atherectomy

Although adjunctive angioplasty after DCA was initially discouraged, angioplasty may actually improve DCA outcome and can often result in better immediate angiographic results. In addition, angioplasty has been shown to convert suboptimal DCA to optimal DCA and improve the clinical outcome. Furthermore, stenting (bare metal) following DCA has resulted in superior luminal enlargement compared to angioplasty alone. In the current era, however, DCA with or without the use of subsequent DES implantation has a minimal role. Nevertheless, it is still occasionally combined with DES when the anatomical setting is appropriate, such as a left main stenosis with a large plaque burden demonstrated by IVUS, and lesion characteristics suitable for removal with current directional atherectomy devices. IVUS plays a further important role in the correct sizing of the device according to the vessel size. This approach, however, has not demonstrated superior efficacy invariably since other studies have suggested that comparable luminal enlargement can be achieved using angioplasty alone. Despite some of these encouraging statements it cannot be denied that the use of coronary atherectomy is in steady decline.

DCA should not be used systematically, but only when the operator needs to achieve significant plaque removal and optimal lumen expansion. Optimal atherectomy has been defined as an angiographic residual stenosis ≤15% [32]. When the operator considers it appropriate to perform DCA, some general recommendations must be followed. First, initial cuts should be directed towards angiographically apparent plaque (as guided by multiple orthogonal views). The goal of optimal atherectomy is to create a large lumen diameter without complications. However, the operoperator should be cautious in upsizing the device because of the risk of perforation. It is worth noting that deep wall components (media and adventitia) can be identified in up to 2/3 of DCA specimens. Although immediate post-procedure lumen diameter is an important determinant of restenosis and is the central theme of the “bigger is better” hypothesis, there is a certain concern that achieving large lumen with partial excision of vessel wall tissue may increase the risk of procedural complications, aneurysm formation and restenosis.

FOCUS BOX 2Performing optimal atherectomy
  • Adjunctive angioplasty after DCA results in better immediate angiographic results and can convert suboptimal DCA into optimal DCA
  • Currently, DCA is almost always followed by stent implantation
  • IVUS plays an important role in the correct sizing of the device according to the vessel size as well as in the characterisation of the plaque
  • DCA works only when the operator uses the device to achieve significant plaque removal and optimal lumen expansion, i.e., optimal atherectomy
  • Optimal atherectomy has been defined as an angiographic residual stenosis ≤15%
  • When performing DCA, initial cuts should be directed towards angiographically apparent plaque (as guided by multiple orthogonal views)

Clinical studies on DCA

The efficacy of DCA in removing non-calcified atherosclerotic plaque has been confirmed in several large-scale randomised trials comparing the results of this device with balloon angioplasty [7, 10, 33, 34]. Angiographic restenosis and target lesion revascularisation rates in the major atherectomy studies are shown in Figure 3 and Figure 4.

The Coronary Angioplasty Versus Excisional Atherectomy Trial (CAVEAT-I) was the first multicentre randomised trial comparing balloon angioplasty with DCA [10]. In this study, 1,012 patients were randomly assigned to either DCA (n=512) or angioplasty (n=500) with the primary endpoint of angiographic restenosis at 6 months. Removing coronary plaque by DCA led to a larger acute gain in lumen diameter (1.05 vs. 0.86 mm; p<0.001) and a small but non-significant reduction in angiographic restenosis (50% vs. 57%; p=0.06) at 6-month follow-up, with no benefit in clinical outcomes [10]. Moreover, DCA was associated with a higher rate of in-hospital myocardial infarction, increased hospitalisation costs, and an excess in mortality at 1-year follow-up (2.2% vs. 0.6%, p = 0.035) [6]. One of the limitations of this study was that the vessel size treated was relatively small and the residual stenosis after DCA relatively high (29%), suggesting that optimal debulking was not achieved. An increase in acute complications with DCA was not confirmed in the Canadian Coronary Atherectomy Trial (CCAT) which compared coronary angioplasty with DCA in 274 patients with non-ostial left anterior descending coronary artery lesions [33]. However, the restenosis rates at 6 months were similar with DCA (46%) and PTCA (43%).

One major limitation of these early studies was the use of a limited (or “non-aggressive”) debulking approach. In the CAVEAT-I study, the larger 7 Fr atherectomy device was used in only 47% of the lesions. However, DCA has subsequently evolved towards a more «optimal» technique (larger devices, more extensive tissue removal, and routine post-dilatation to obtain a final diameter stenosis <20%).

The Optimal Atherectomy Restenosis Study (OARS) was a registry which showed that the use of an “optimal” atherectomy technique (using IVUS guidance and adjunctive balloon post-dilatation, if necessary) to produce larger acute lumen diameters (increasing MLD from 1.18 mm to 3.16 mm, reducing diameter stenosis from 64% to 7%) translated to a lower restenosis rate at 6 months (28.9%) than previously seen with this device, without an increase in early or late major adverse events (2.5% and 23.6% respectively) [32]. Optimal DCA, defined as <15% residual stenosis (by the angiographic core laboratory), with adjunctive balloon dilation if necessary, was achieved in 82% of lesions.

These favourable results led to the Balloon vs. Optimal Atherectomy Trial (BOAT), which evaluated the short-term and long-term clinical outcome of optimal DCA compared with balloon angioplasty in 1,000 patients with single, de novo, native vessel coronary disease. In this study, the encouraging results of the OARS registry were confirmed by the reduced angiographic restenosis rates with atherectomy compared to balloon angioplasty (31.4% vs. 39.8% respectively, p=0.016), without an increase in early adverse events (death, Q-wave myocardial infarction or emergent coronary artery bypass graft surgery; 2.8% vs. 3.3% respectively, p= 0.7) [7]. However, this study did not show a difference in 6-month clinical outcomes (death, Q-wave myocardial infarction or target vessel revascularisation): 21.1% vs. 24.8% respectively, p= 0.1.
In the OARS trial, despite “optimal directional coronary atherectomy” by angiographic criteria, the residual plaque burden assessed by IVUS remained high (58%). This led to the Adjunctive Balloon Angioplasty after Coronary Atherectomy Study (ABACAS), an IVUS-guided, aggressive directional atherectomy study in 214 patients comparing balloon post-dilatation versus no adjunctive balloon post-dilatation. ABACAS showed a further decrease in the 6-month angiographic restenosis rate in both arms (23.6% vs. 19.6% respectively, p= NS) as well as in the target lesion revascularisation rate (20.6% vs. 15.2% respectively, p= NS) [35].

Finally, the STent versus directional coronary Atherectomy Randomised Trial (START) was designed to compare bare metal stenting (62 lesions) to optimal IVUS-guided DCA (60 lesions) in vessels suitable for both techniques [36]. Although the post-procedural lumen diameters were similar (2.79 mm vs. 2.90 mm respectively), the follow-up MLD was significantly smaller in the stent arm (1.89 mm vs. 2.18 mm; p=0.02). IVUS revealed that intimal proliferation was significantly larger in the stent arm than in the DCA arm (3.1 mm2 vs. 1.1 mm2; p<0.0001), accounting for the significantly smaller follow-up lumen area in the stent arm (5.3 mm2 vs. 7.0 mm2, p=0.03). Restenosis was significantly lower (32.8% vs. 15.8%, p=0.032), and target vessel failure at 1 year tended to be lower with DCA compared to stenting (33.9% vs. 18.3%, p=0.056). The authors of this study concluded that aggressive DCA might provide superior angiographic and clinical outcomes compared to primary stenting.

Despite some encouraging results, DCA as a stand-alone technique is more complex and gives less consistent angiographic results than stenting. These limitations fostered the growth of a combined strategy of DCA plus stenting [5, 11, 12, 13, 14, 15]. Kiesz et al were among the pioneers to postulate that debulking before stenting would facilitate complete stent expansion and further decrease elastic recoil at the treated site without the necessity for aggressive stent post-dilation [11]. In the Acute Directional coronary Atherectomy Prior To Stenting in complex coronary lesions study (ADAPTS), DCA plus stenting was performed in 60 consecutive patients with 89 lesions considered high-risk for restenosis (aorto-ostial lesions, chronic total occlusions, long lesions, and lesions containing thrombus) and the restenosis rate was 13.3% [11]. Similar restenosis rates after DCA and stenting were reported by Bramucci et al (6.8%) [13], Hopp et al (AtheroLink Registry, 8.4%) [14], and Airoldi et al (13.8%) [37]. In the Stenting after Optimal Lesion Debulking (SOLD) Registry, 128 patients with 168 complex lesions were enrolled and restenosis rate was 11% and the target vessel revascularisation rate was 7% [5].

The encouraging results of DCA and stenting seen in registries led to the initiation of two large, prospective, randomised clinical trials comparing DCA before stenting with stenting alone: the Atherectomy Before MULTI-LINK Improves Lumen Gain and Clinical Outcomes (AMIGO) trial [38] and the Debulking and Stenting In Restenosis Elimination (DESIRE) trial [4]. The AMIGO trial randomised 753 patients with de novo or restenotic lesions to either DCA followed by stenting or stenting alone [38]. There was no increase in 30-day major adverse cardiac events in the DCA plus stent group (3.9% vs. 2.4%, p=0.30). The primary endpoint, angiographic restenosis at 8 months, occurred in 26.7% of patients treated with DCA plus stents and in 22.1% of patients treated with stents alone (p=0.237). However, this study failed to support the original findings and hypothesis. Optimal debulking (defined as post-DCA diameter stenosis <25%) was achieved in only 26.5% of patients despite the fact that the study protocol required this endpoint in all patients who were randomised to the DCA arm. Suboptimal debulking was associated with a significantly higher restenosis rate (32%) compared with optimal debulking (16%; p=0.01).

The DESIRE trial randomised 500 patients to IVUS-guided DCA followed by stenting or stenting alone. Despite the achievement of a lower loss index at quantitative coronary angiography follow-up in the DCA+stent group (0.34 vs. 0.41, p=0.05), this did not translate to clinical benefit at 6-month follow-up. The main problem of DCA, which may also have affected the results of the AMIGO trial, is that the technique is very operator-dependent and the amount of tissue removal varies depending on the commitment of the operator to perform extensive debulking.

A meta-analysis performed by Niccoli et al involving 12 randomised and non-randomised trials comparing DCA before stenting versus stenting alone, showed that debulking was superior as regards angiographic results and revascularisation rates with no difference in major adverse events [39]. This analysis included 1,216 patients with DCA plus stenting and 1,484 undergoing stenting alone. Debulking yielded greater acute lumen gain when compared to stenting alone, and was associated with significantly lower rates of angiographic restenosis (OR 0.67, 95% CI 0.54–0.84; p=0.0003) and target lesion revascularisation (OR 0.73, 0.59–0.91; p=0.006).

The widespread adoption of DES and their efficacy in reducing restenosis has resulted in DCA falling out of favour and being relegated to specialist centres that continue to believe in the utility of this technology. Indeed, the published data of DCA followed by DES implantation is quite limited. One such study is the PERFECT (PrE Rapamycin-eluting stent FlExi-CuT) registry [40, 41], which does however suggest that such niche technology as DCA only has a favourable cost-benefit ratio when used selectively and appropriately. In the PERFECT prospective multicentre registry, 99 patients underwent IVUS-guided DCA of coronary bifurcation lesions prior to DES implantation. Eighty-one per cent of the lesions were located in the left main or at the ostium of the LAD or LCX. Atherectomy was performed predominantly in the MB, with only 3 lesions treated in the side branch as well. No complications occurred during the procedure. The primary endpoint of binary restenosis, occurred with 1 lesion (1.1%) in the main branch and with 3 lesions (3.4%) in the side branch, giving a total restenosis rate of 4.5%. Target lesion revascularisation was performed in 2 patients (1 for the main branch and the other for the side branch). No deaths, coronary artery bypass grafting or myocardial infarctions were reported within the first year. In the 63 patients with lesions located in the left main, there were no cases of angiographic restenosis.

In a more recent observational study, Tanaka et al compared the outcomes of debulking with DCA (n = 41) versus no DCA (n = 60) prior to implantation of a single DES in patients with unprotected left main bifurcation lesions. Plaque debulking with DCA was associated with a significantly reduced plaque burden at the LCX ostium (percent diameter stenosis after PCI: 19.2±13.1% vs. 28.3±22.7%, p=0.034; at follow-up: 20.8±12.3% vs. 31.9±21.4%, p=0.007) as well as reduced restenosis at the LCX ostium at 9-month angiographic follow-up (0% vs. 10.2%; p = 0.048) [42].

A novel recent approach with DCA in Japan has been to utilize a drug-coated balloon after DCA of bifurcations in order to avoid stenting. This approach could potentially be very appealing in ostial circumflex lesions where the restenosis with DES still remains very high. Kitani et al treated 129 patients with this combined DCB and DCA strategy with a TLR rate of only 3.1% at 12-months [67].

In summary, the available evidence indicates that DCA does not improve late angiographic outcome when performed before bare metal stent implantation unless optimal debulking is achieved and relatively higher risk lesions are treated. To support this second statement only indirect data exists. The utility of plaque debulking with DCA prior to DES implantation has not been well tested but the 2 registry studies presented above do suggest that a re-evaluation of the role of DCA in complex lesions such as the left main bifurcation is warranted.

FOCUS BOX 3Clinical studies on DCA
  • CAVEAT-I and CCAT compared DCA to balloon angioplasty and both utilised a non-aggressive debulking approach
  • OARS, BOAT, ABACAS and START were registries using IVUS-guided optimal DCA, which resulted in lower restenosis rates than those in CAVEAT-I and CCAT
  • SOLD, ADAPTS, AtheroLink and ADAPTS were the first registries to evaluate DCA followed by bare metal stenting which demonstrated encouragingly low restenosis results and led to the initiation of two randomised trials (AMIGO and DESIRE)
  • AMIGO randomised 753 patients with de novo or restenotic lesions to DCA+stenting vs. stenting alone. No significant difference in restenosis and optimal debulking was performed in only 26.5% of patients
  • DESIRE randomised 500 patients to DCA+stenting vs. stenting alone and also failed to show a clinical benefit from DCA
  • A meta-analysis of these trials showed that DCA was associated with significantly lower restenosis and revascularisation rates
  • There are only 2 registry studies evaluating DCA debulking prior to DES implantation which both show very promising results in left main bifurcation lesions
  • In summary, the available evidence indicates that DCA does not improve late angiographic outcome when performed before bare metal stent implantation unless optimal debulking is achieved and relatively higher risk lesions are treated

Avoiding procedural complications with DCA

DISSECTION

The rates of non-occlusive dissection and severe dissection leading to abrupt closure, with the early DCA devices, were 20% and 0-7% of cases respectively [43]. In CAVEAT-I, abrupt closure was more common in the DCA compared to angioplasty (8% vs. 3.8%, p=0.005) and occurred at a site other than the target lesion in 42% [34]. Dissections can be caused by the guide catheter, the guidewire, and the atherectomy device itself (from the cutting mechanism, the nose cone, or more rarely from the integrated balloon in the case of the Flexi-Cut™ device). Avoidance of deep seating and over-rotation of the guiding catheter can minimise the catheter-induced injury. It is always important not to overinflate the urging balloon. One to a maximum 3 atmospheres is frequently all that is needed to establish appropriate plaque contact. Avoiding atherectomy at extremely angulated segments will also minimise the risk of dissections and vessel perforation.

DISTAL EMBOLISATION AND THE NO-REFLOW PHENOMENON

This complication occurred in 0-13.4% of the DCA cases in the earlier studies [34, 43, 44, 45, 46, 47]. This type of complication may occur due to dislodgement of thrombus or soft plaque from the target lesion, or incomplete capture of material from the cone of the device [45]. No-reflow is more frequent in vein grafts and thrombus containing lesions. There are numerous proposed treatments for this complication, which invariably include the intracoronary administration of a vasodilator such as nitroprusside, verapamil, or adenosine.

NON-Q-WAVE MYOCARDIAL INFARCTION AND SIDE BRANCH OCCLUSION

Side branches may be compromised at a rate of 0.7% to 7.7% after DCA [32, 34, 43, 45, 46, 47, 48]. Careful selection of cases, such as avoidance of DCA in the treatment of bifurcations with side branches with a diseased origin, can help in reducing the rate of this complication unless the operator is able to perform initial cuts towards the origin of the side branch. This condition is a typical situation where experience plays a fundamental role: paradoxically, a condition where the side branch could become totally occluded can be converted into the ideal setting for atherectomy if handled appropriately. Indeed, DCA has been used successfully to salvage side branches in bifurcation lesions with significant plaque at the ostium of the side branch and the impossibility of wiring the side branch. Correctly performed debulking can result in the side branch becoming accessible.
DCA was implicated in CAVEAT-I as being associated with a higher incidence of non-Q-wave myocardial infarction compared to traditional angioplasty [6]. However, these results were not reproduced in the CAVEAT-II and CCAT trials [6, 33, 49]. Risk factors for CK-MB release were complex lesion morphology and de-novo lesions.

PERFORATION

A coronary artery perforation occurred in fewer than 1% of cases in early DCA studies [6, 11, 32, 34, 49, 50]. With the newer devices (i.e., SilverHawk™) and with the use of standard workhorse wires this complication is rare. Careful checking of the distal end of the wire is important to prevent wire perforations. Free mobility of the distal guidewire should be maintained at all times. Loss of free mobility may suggest that the collection chamber is full. As we suggested previously, it is important to avoid cutting in extremely angulated segments especially when using the Flexi-Cut™ device which has a longer straight housing compared to the SilverHawk™.

VASOSPASM

Severe vasospasm occurred in fewer than 2% of cases with the use of the early DCA devices [2, 34, 47, 51]. Spasm will respond readily to intracoronary nitroglycerine or other vasodilators or even gentle low-pressure balloon inflations.

THROMBOSIS

Thrombosis occurred in approximately 2% of cases and accounted for half of the abrupt closure cases in the early studies of DCA [2, 34, 47, 51]. With the optimisation of antiplatelet therapy and especially with the addition of clopidogrel to aspirin, this problem seems to have been reduced dramatically. In the absence of other procedural complications, strict adherence to optimal dual antiplatelet pre-treatment will minimise the risk of intraprocedural thrombosis. The usage of glycoprotein IIb/IIIa inhibitors is certainly a reasonable idea to lower the risk of this complication. In our experience we have kept the usage of glycoprotein IIb/IIIa inhibitors quite low and rarely administer them on an elective basis (less than 10%) due to the consequences if a perforation occurs.

Associated pharmacological treatment

Periprocedural preparation with thienopyridines and the duration of combined thienopyridine and aspirin treatment following stent implantation after DCA should follow contemporary ESC guidance.

When performing DCA with stenting, a standard protocol of periprocedural heparin administration (100 U/Kg withou IIb/IIIa and 70 U/Kg with IIb/IIIa) should be used. A glycoprotein IIb/IIIa inhibitor is sometimes administered when the final result appears suboptimal and when, for various clinical or anatomical reasons, the operator feels wants to avoid implanting a stent. In addition, the use of a glycoprotein IIb/IIIa inhibitor prior to DCA has been shown to reduce the incidence of non-Q-wave myocardial infarction [52]. As already mentioned, however, the elective usage of IIb/IIIa inhibitors is low in clinical practice due to the potential detrimental consequences of a coronary perforation.

FOCUS BOX 4Avoiding procedural complications with DCA
  • Avoid dissections by careful guiding catheter manipulation without deep seating or over-rotation, not overinflating the urging balloon, and not performing cuts on angulated segments
  • No-reflow usually occurs when performing DCA in vein grafts or thrombus-containing lesions and can usually be resolved with intracoronary vasodilators
  • Side branches may become occluded during DCA but correctly performed DCA at a bifurcation can also be used to salvage side branches
  • Coronary perforations can occur due to cuts on extremely angulated segments or from the distal guidewire. Although this complication is rare, it is imperative that PTFE-covered stents are available when planning to perform DCA

The role of DCA in current interventional practice

Despite the lack of scientific evidence supporting plaque debulking in today’s clinical practice, experience when the anatomical setting is appropriate has been favourable. Atherectomy and DES are thus still occasionally combined, such as for specific ostial, bifurcation, and left main lesions. Despite this statement, it should be recognised that the use of atherectomy, even in suitable lesions, continues to decrease. An additional important limitation is the fact that presently there are no approved DCA devices in Europe or in the USA which can be used in the coronary vessels. DCA can thus only be performed with a peripheral device as an off-label indication.

OSTIAL LESIONS

Percutaneous interventions on ostial lesions, particularly aorto-ostial lesions, are frequently limited by lesion rigidity and elastic recoil, leading to suboptimal results [53, 54, 55]. For non-calcified ostial lesions in vessels >3 mm, DCA is associated with procedural success in 87% and major complications in <1% of patients [53, 54]. Although immediate angiographic results in highly selected lesions are excellent, DCA of ostial lesions is limited by a high incidence of restenosis. It has previously been shown that this problem seems to be eliminated by the implantation of DES with or without prior DCA in the setting of aorto-ostial lesions [56].

BIFURCATION LESIONS

Angioplasty on bifurcation lesions is often complicated by the “snowplough” effect or ‘’plaque shift’’ which can be complicated by side branch occlusion, and suboptimal final angiographic results [57, 58]. These problems continue, although to a lesser extent, with the treatment of bifurcations with DES. The risk of side branch occlusion with DCA is greatest when the side branch originates from the target lesion, or if it is significantly diseased (stenosis >50%). In contemporary practice it is occasionally possible to treat the main branch with DCA, when there is a large plaque burden, in order to avoid plaque shift.

LEFT MAIN LESIONS

Stenoses in the left main coronary artery are well-suited for DCA because of their proximal location and large vessel calibre. It is still o ccasionally combined with atherectomy and DES when treating a left main stenosis with a large plaque burden, demonstrated by IVUS to have the characteristics suitable for removal with DCA [59, 60].

IN-STENT RESTENOSIS

Even in the era of DES, ISR remains a challenging problem with a high rate of recurrence after retreatment [61]. DCA may be advantageous in selected cases with marked in-stent neointimal proliferation, as atherectomy facilitates neointimal tissue resection, thus allowing for more optimal repeat revascularisation. Registry studies have suggested that DCA for bare metal stent restenosis results in large post-procedural lumen diameters and low rates of target lesion revascularisation, and that it is possibly superior to rotational atherectomy [62, 63]. Finally, DCA allows for characterisation of in vivo tissue histopathology and helps elucidate the mechanisms of restenosis [64].

FOCUS BOX 5Favourable anatomical settings for DCA
  • Ostial lesions
  • Bifurcation lesions
  • Left main lesions
  • In-stent restenosis

Conclusions

The survival of debulking as a percutaneous approach depends on the demonstration of a benefit in terms of restenosis and long-term outcomes when combined with stenting. The superior anti-restenotic efficacy of DES and the results of the AMIGO and DESIRE trials cast doubt on the current role of atherectomy. Despite these concerns, in selected complex lesions (e.g., bifurcations, ostial lesions), no device including DES is likely to be infallible. It is possible that in these settings plaque removal prior to stent implantation will allow greater and more uniform stent expansion and minimise plaque shift, thereby improving angiographic and clinical results. However, in order for this concept to become a practical reality, a further refinement of atherectomy devices with increased efficiency and safety and lower profile will be necessary. In the era of DES, the clinical (and/or market) needs may not be great enough to generate sufficient industry commitment and investment to allow these developments to occur, nor for their utility to be demonstrated in appropriately designed clinical trials. Thus, although DCA is currently little more than a niche device, it remains in the armamentarium of interventional cardiologists treating complex lesions, particularly complex left main bifurcation lesions with a large plaque burden.

Personal Perspective - Antonio Colombo

Procedural complexity and long-term recurrence remain major concerns when stents are implanted in “complex lesion subsets”, such as long lesions, ostial lesions, bifurcations, and calcified or non-dilatable lesions. Pre-treatment of these complex lesions with high-pressure balloon inflation before stent implantation is certainly an option, but it is not always successful. Suboptimal dilatation, acute recoil, plaque shift, dissections and vessel perforation are all potential shortcomings with this approach. To remedy this problem, DCA in non-calcified lesions and rotational atherectomy in calcified lesions have been developed to prepare complex plaques before stent implantation. However, the encouraging results of many single-centre experiences [65] was not reproduced in the context of randomised studies [4, 13]. The technique is very operator-dependent and the amount of tissue removal varies depending on the commitment of the operator to perform extensive debulking. In addition, apart from the very recent introduction of the SilverHawk™ device, there has been no further development in the available devices for some considerable time. The introduction of DES has further challenged the utilisation of additional complex approaches and devices. Despite these concerns and the lack of convincing scientific evidence supporting the advantage of plaque debulking in bifurcation lesions, our experience in this setting has been favourable and we still occasionally combine atherectomy and DES when the anatomical setting is appropriate, such as a left main stenosis with a large plaque burden. The rationale of plaque debulking with atherectomy is as appealing as ever and the combination of DCA with DES remains a logical combination which is still waiting to be substantiated.

Online data supplement

ILLUSTRATIVE CASE 1
Case illustration of lesion involving the ostium of the left anterior descending artery (panel A) treated with atherectomy and stenting (panel B), and minimal late loss at 8-month follow-up (panel C).
ILLUSTRATIVE CASE 2
Case illustration of a lesion involving the ostium of the left circumflex coronary artery (panel A) treated with atherectomy with the tip of the atherectomy catheter across the lesion during cutting (panel B), and final angiographic result following stenting (panel C).
ILLUSTRATIVE CASE 3
Case illustration of a true bifurcation lesion involving the left anterior descending artery and diagonal branch (panel A), treated with atherectomy (panel B), followed by stenting on both branches of the bifurcation (panel C).

SHARE YOUR COMMENT

SHOWING 1 COMMENT