Abstract
Mechanical debulking of coronary plaques with rotational atherectomy (RA) has been used for more than 20 years during percutaneous coronary interventions (PCI). Modification of plaque characteristics may be accomplished with selective ablation of inelastic fibrocalcific tissue. The use of RA, though reduced with the development of bare-metal stents (BMS) and even more with drug-eluting stents (DES), has never been completely abandoned. The present review will analyze reasons for conflicting results obtained in large series and randomized trials on this topic in the past, and will identify criteria for an appropriate use in current times.
1
Introduction
In the late ’80s, in order to increase the efficacy of balloon angioplasty, some intravascular devices were designed with the aim to debulk the atherosclerotic plaque and increase the likelihood of obtaining a larger lumen, according to the concept of “bigger is better”.
A high-speed rotational atherectomy (RA) device was preliminarily tested in cadavers under angioscopic guidance by David Auth ; the system was subsequently mounted on a flexible rotating shaft and used in human coronary arteries with a standard angioplasty equipment in 1988 .
Since then, debulking has been used in the treatment of complex coronary lesions, sometimes as a stand-alone technique, later as preparatory to balloon dilation and more recently to stent deployment.
RA utilization peaked in mid ’90s, when mainly high-volume French centers performed RA-assisted percutaneous coronary intervention (PCI) in up to 20% of cases , then its use progressively decreased with the advent of stent-assisted PCI . The volume of RA has varied widely through these two decades and between individual centers—from 0.6% to 8.2% among centers adopting this strategy —but its use has never been completely abandoned in daily practice, even after the advent of drug-eluting stents (DES); from national registries it is evident that slightly less than 1% of patients submitted to PCI are treated with RA annually ( Fig. 1 ) .
Other plaque ablation techniques were developed in the same period—directional coronary atherectomy (DCA), transluminal extractional atherectomy catheter (TEC), among others —but their use has been almost abandoned. Only excimer laser coronary angioplasty (ELCA) survived, and is currently being tested as a thrombus dissolving device in the setting of ST-elevation myocardial infarction (STEMI) .
The aim of the present manuscript is to review the conflicting evidence that induced a progressive reduction in the applications of the technique and also to identify subsets of patients who might still hypothetically benefit from RA in the current DES era.
2
Methods
PubMed and EMBASE databases were searched for articles written in English and reporting percutaneous coronary intervention performed with rotational atherectomy. The time frame for the articles searched was January 1991 through December 2011. The following search terms were utilized: “rotat*” and “atherect*” and “coronary”. References of the articles identified in this manner were also searched through to locate additional references that—not identified by the search strategy—might be useful for the purpose. Clinical studies, which included randomized trials, prospective cohort studies, and retrospective analyses, were included; abstracts and case reports were excluded. Discordance regarding inclusion was resolved by discussion.
2.1
Rationale of rotational atherectomy
RA is accomplished with the Rotablator (Boston Scientific-Scimed Corporation, Natick, Massachusetts), equipped with a diamond chips-coated spinning burr that removes plaque by abrading the atherosclerotic material, producing particles that should be < 10 μm; such microparticles are usually dispersed into the distal coronary circulation and cleared by reticulo-endothelial system in liver, lung, and spleen. However, these particles may have a detrimental effect on the myocardial microcirculation and a higher risk of periprocedural myocardial infarction (MI) has been described after RA-assisted PCI .
The two major principles that govern the ability of the Rotablator system to treat atherosclerotic lesions are “differential cutting” and “orthogonal displacement of friction” . “Differential cutting” is defined as the ability to selectively ablate one material while maintaining the integrity of another that can be deflected away from the advancing rotating abrasive burr: inelastic, hard and even calcified atherosclerotic plaque is abraded while deflecting the normal, soft and elastic tissue. At rotational speeds > 60,000 rpm the friction, which occurs when sliding surfaces are in contact, is virtually eliminated. “Orthogonal displacement of friction” refers to the change in the effective longitudinal friction—which is almost eliminated—resulting in reduced surface drag and unimpeded advancement of the burr in tortuous and diseased segments of the coronary tree .
2.2
Technique
The Rotablator system includes an advancer that houses the air turbine, drive shaft, and burr and a console to monitor the rotation by regulating air supply to the advancer and the Dynaglide foot pedal. The abrasive tip is welded to a long flexible drive shaft tracking along a central flexible stainless steel guidewire. The drive shaft is housed in a 4.3 F Teflon sheath which prevents tissue injury caused by the spinning shaft and also acts as a conduit for a flush solution; the system may be flushed either with the Rotaglide Lubricant—a phospholipid solution that reduces heat and friction during burr advancement—or with saline plus heparin and various vasodilating drugs (nitroderivates and verapamil) in order to reduce vasospasm. The abrasive tip is an elliptically shaped burr, available in various sizes for coronary use (1.25–2.50 mm in diameter). The distal half is coated with diamond chips 20–30 μm in diameter. Rotational energy is transmitted by a compressed air motor that drives the flexible helical shaft at speeds up to 200,000 rpm. The number of rpm is measured by a fiber optic light probe and displayed on a console. The speeds of rotation and of advancement of the burr are controlled by the operator. During rotation, saline-based flush solution irrigates the catheter sheath to lubricate and cool the rotating parts. The burr and the drive shaft move freely over a central coaxial guidewire, the RotaWire (0.009” diameter, 300 cm in length), with a flexible radiopaque platinum distal tip (0.014” in diameter, 20 mm long), which is locked and does not rotate with the burr. The RotaWire is available in two designs: Floppy, with a long tapered shaft, for a greater flexibility to negotiate tortuous vessels and Extra Support, with a short tapered shaft, to maximize vessel straightening. The RotaWire and the abrasive tip can be advanced independently, which allows the wire to be placed in a safe distal location before the burr is advanced into the diseased artery .
In RA, the selection and the positioning of the guide catheter are important for the ease of advancement of the burr. Coaxial alignment of the guide catheter and guidewire has to be verified before ablation. The inner diameter of the guiding catheter has to be 0.004” larger than the final burr. Most 8 Fr guiding catheters permit passage of up to a 2.15 mm device; a 7 Fr may accommodate the 2.0 mm, while up to a 1.75 mm burr can be advanced in a 6 Fr guiding catheter.
Very complex lesions can be first crossed with a regular coronary guidewire, with subsequent exchange with an RA guidewire through an over-the-wire balloon catheter or a microcatheter.
Among RA treatment strategies, the Rotablator stand-alone technique has been early abandoned, as even aggressive burr sizing (burr/artery ratio in the range 0.7–0.9) did not allow satisfactory luminal gain and was burdened with an unacceptable high in-hospital complication rate (> 10%), as compared to intermediate burr sizing in the Coronary Angioplasty and Rotablator Atherectomy Trial (CARAT) . Reisman et al. demonstrated that excessive drops in speed and aggressive advancement of the burr were related to significant increases in temperature and potential thermal injury. Indeed, clinical studies have corroborated these experimental findings: in the randomized Study to Determine Rotablator and Transluminal Angioplasty Strategy (STRATAS) trial , decelerations > 5000 rpm from baseline for a cumulative time > 5 s were associated with both an increase in CK-MB elevation and restenosis.
Currently “optimal” RA technique includes a slow advancement of the burr, with a to-and-fro pecking motion, short run times (15–20 s) at “low” speed (140,000 –160,000 rpm); care must be directed to avoid any “significant” (> 5000 rpm) drop in burr speed. Nowadays, in order to minimize risks, use of multiple RA burrs (the “step-up” technique) has been abandoned, as well as the concept of RA as a “stand-alone” technique. RA is used as a “facilitation” for subsequent balloon or stent expansion, with a single burr selected with a burr/artery ratio in the range of 0.5–0.6; a 0.7 ratio can be used when treating a segment already “protected” by a stent previously deployed, as in the case of in-stent restenosis (ISR).
2.3
Complications
The use of the Rotablator is technically demanding, and associated with a significant incidence of adverse events.
There are substantial differences in the complication rates among the various series, and this is due to the heterogeneity in the definitions used, the variable operator experience with the device, the evolution of the technique and the choice of lesion and patient subsets . Larger series and clinical trials report death in 1%–5% of cases, myocardial injury as assessed as a Q-wave MI in 0.7%–4.8% or an increase of CK-MB in 0.6%–18.7%, and urgent coronary artery bypass grafting (CABG) in 0.4%–2.5% .
Procedural complications include abrupt occlusion in 1.1%–11.2% of cases, whose primary causes are coronary dissection, less frequently spasm and side-branch closure. The risk of dissection is predicted by the presence of vessel tortuosity, lesion complexity > B type and primary rather than restenotic lesion; in pre-stent era coronary dissection was the main determinant of urgent CABG .
Severe spasm is extremely frequent after burr passage; the incidence of spasm has been contained with intracoronary continuous infusion of nitroglycerin and nicorandil .
The no-reflow phenomenon refers to a state of myocardial tissue ischemia due to reduced coronary flow in the presence of a patent epicardial coronary artery. The underlying cause of no reflow is microvascular obstruction, which may be produced by various mechanisms, being distal embolization of thrombus and atherosclerotic fragments the most likely culprits . It is documented in > 20% of patients undergoing primary PCI for STEMI and in < 2% of elective PCI cases . The peculiar mechanism of action of RA—plaque abrasion with microparticle embolization—leads to an increased incidence of no-reflow, with vasoconstriction and/or mechanical obstruction of distal coronary resistance vessels. Intracoronary administration of powerful arteriolar vasodilators, such as nicardipine, adenosine and nicorandil—more than verapamil—showed a protective effect on the occurrence of no-reflow .
PCI of bifurcating lesions is weighed with the plaque or carina shift (the “snow-plow” effect) that can cause the occlusion of the side branch. The “differential cutting” should theoretically allow RA to accomplish a selective abrasion of plaque in the proximity of a branch, thus increasing the procedural success rate and reducing the need for side-branch intervention .
Compared to balloon PCI, RA has been associated with a 4-time increased risk of perforation : in large series of unselected patients undergoing PCI it has been reported in 0.48% . Coronary perforation can be sealed with prolonged balloon inflation or stent deployment—more likely covered; however, even in more recent series of patients experiencing a perforation, emergency CABG has been reported in 2.9%, periprocedural MI in 7.4% and in-hospital mortality rate in 5.9% of cases .
Bradyarrhythmias occur frequently during RA, but can be prevented with short durations of burr activation and advancement and are easily reversed with intravenous atropine; when treating a severely calcified large-sized right coronary artery the preventive placement of a temporary pace-maker should be considered.
The risk of significant bleeding and other vascular complications related to large sheath size has been reported in the range of 1.0%–5.0%, and is probably lower today, since large burrs requiring sheaths > 8 Fr are rarely used.
2.4
Adjunctive pharmacology
High-speed RA increases the risk of intravascular thrombus formation as compared to traditional PCI: using ex-vivo models, a direct relationship was documented between the degree of burr speeds, heat generation and increased platelet activation ; subsequent enhanced platelet aggregation has been reduced with glycoprotein (GP) IIb-IIIa inhibitors’ pretreatment . Aggressive antiplatelet therapy is therefore mandatory when RA is planned.
GP IIb-IIIa inhibitors have been extensively used in clinical practice: abciximab—the most commonly used—was shown to reduce the risk of periprocedural myocardial injury as assessed by either myocardial perfusion imaging or creatine kinase-MB (CK-MB) elevation . However, such benefit was mainly reported before the advent of systematic high-dose clopidogrel pre-treatment, and was not confirmed in later large-scale registries .
Bivalirudin is an attractive alternative to heparin + GP IIb-IIIa inhibitors in RA-assisted PCI, due to its reduced hemorrhagic risk profile; a limited experience is available, aiming to document only the safety profile of direct antithrombin inhibition .
There are no data currently available on the efficacy of newer P2Y12 blocking agents (prasugrel and ticagrelor) in patients undergoing RA-assisted PCI.
A protective effect of pre-RA statin treatment has been documented, with a reduction in periprocedural myocardial injury as compared to patients not taking lipid-lowering drugs .
Due to the frequent occurrence of coronary spasm and microvascular obstruction, vasodilators, such as calcium channel blockers (verapamil, diltiazem, or nicardipine), adenosine, nicorandil or nitroprusside should be pre-administered and liberally used during RA in order to reduce the occurrence of the “no-reflow” phenomenon .
2.5
Clinical applications of rotational atherectomy over time
Unfortunately, clinical trials testing the effectiveness of RA-assisted PCI in native coronary artery disease (CAD) as compared to balloon PCI failed to document improved outcomes. In both Excimer Laser, Rotational Atherectomy, and Balloon Angioplasty Comparison (ERBAC) and COmparison of Balloon angioplasty versus Rotational Atherectomy (COBRA) trial, as compared to stand-alone balloon PCI, RA-assisted balloon PCI obtained similar or at best slightly higher procedural success rates, but late results were disappointing, with similar 6-month target lesion revascularization (TLR) rates ( Table 1 ). However, lesion complexity was the only morphological selection criterion in these studies; the Dilatation vs Ablation Revascularization Trial (DART) enrolled only patients with lesions located in small vessels (diameter < 3 mm), and the presence of calcification was even an exclusion criterion. Calcified and ostial or bifurcating lesions are the most likely substrates where debulking can alter plaque morphology, making the plaque more distensible and allowing a larger lumen gain with subsequent balloon expansion ( Fig. 2 ). In randomized trials, results of RA were analyzed in “complex” lesions as a whole—being such group heterogenous, as including long, eccentric, tortuous, occluded lesions—and positive results that might have been obtained only in calcified and ostial lesions were not demonstrated. Moreover, pooling data from randomized trials of balloon PCI vs atheroablative techniques (coronary atherectomy, laser angioplasty, or cutting balloon atherotomy) a meta-analysis documented that ablation increases the risk of periprocedural MI, without any benefit in terms of short- and 6-month survival nor a reduction in the restenosis rate .
| Year | Study design | Main Results (RA-assisted vs “other” PCI) | |
|---|---|---|---|
| RA vs balloon PCI studies | |||
| ERBAC ( n = 685) | 1997 | RA vs ELCA vs balloon PCI | Procedural success 89% vs 77% vs 80% ( P < 0.01); 6-month TLR 42% vs 46% vs 32% ( P < 0.05). |
| COBRA ( n = 502) | 2000 | RA vs balloon PCI | Procedural success 84% vs. 73% ( P < 0.05); similar 6-month TLR. |
| DART ( n = 446) | 2003 | RA vs balloon PCI in < 3.0 mm vessels | Similar procedural success, binary 8-month restenosis and 12-month adverse events. |
| RA strategy studies | |||
| STRATAS ( n = 500) | 2001 | B/A 0.7–0.9 with < 1 atm balloon PCI vs B/A 0.6–0.8 with > 4 atm balloon PCI | Similar procedural success; 6-month TLR 31% vs 22% ( P < 0.05) |
| CARAT ( n = 222) | 2001 | B/A > 0.7 vs B/A ≤ 0.7 | Angiographic complications 13% vs 5% ( P < 0.05); similar procedural success and 6-month TLR. |
| RA-assisted vs balloon-assisted DES studies | |||
| DOCTORS ( n = 266) | 2008 | RA or DCA-assisted vs balloon predilation DES deployment in CTO | 30-day MACE rate 16% vs 8.5% ( P = 0.07), 1-year MACE 27.5% vs 40% ( P = 0.033). |
| ROTAXUS ( n = 240) | 2011 | RA-assisted vs balloon predilation DES deployment in calcified lesions. | Superior acute gain ( P = 0.03) counterbalanced by a higher late loss ( P < 0.01) yielded similar restenosis rate. |
Later on, in the ’90s, when stent became state-of-the-art PCI procedure, its optimal deployment with symmetrical strut expansion was generally pursued with high pressure balloon inflations; again, in calcified lesions this was not systematically obtained and RA-assisted stent deployment has been proposed in selected cases as the “rotastent” technique . In theory, residual plaque burden left outside the stent being directly proportional to the amount of neointimal proliferation , mechanical debulking would have been beneficial to reduce recurrences by decreasing the amount of underlying plaque burden before stent deployment.
Unlike randomized studies, registry data showed the effectiveness of RA-assisted PCI in calcified and ostial/bifurcating lesions ; in order to overcome the absence of true comparison with standard balloon predilation, a propensity score analysis would have been conducted on large data, and only this analysis would have allowed a robust support to the adequacy of choice for operators when performing RA in complex lesions. No randomized trial was ever conducted to evaluate the benefit of RA-assisted stent-PCI as compared to a standard stent deployment technique in severely calcified lesions.
Bare metal stent (BMS) implantation was plagued with neointimal proliferation inside the struts, and the treatment of ISR gave disappointing results, with a rate of recurrences associated with the angiographic pattern of ISR: TLR was 19% for focal (≤ 10 mm) disease (Mehran pattern I), 35% for neointimal hyperplasia > 10 mm within the stent (pattern II), 50% for ISR > 10 mm extending outside the stent (pattern III), and 83% for totally occluded ISR (pattern IV) . After the documentation that in a porcine model neointimal ablation with RA was associated with significantly less recurrent neointimal hyperplasia than after balloon PCI , Rotablator was advocated for the treatment of diffuse ISR (Mehran’s pattern III and IV) but randomized trials were once again inconclusive ( Table 2 ). As assessed by intravascular ultrasound (IVUS), RA reduced neointimal hyperplasia volume (43 ± 14) more significantly than ELCA (19 ± 10 mm 3 , P < 0.001); this aggressive treatment did not translate into clinical benefit, as procedural success rate, in-hospital outcome and 6-month TLR were similar with both atheroablative strategies . Trials designed to compare RA-assisted balloon PCI vs balloon PCI for the treatment of ISR gave again controversial results: in the Angioplasty versus Rotational atherectomy for Treatment of diffuse In-Stent restenosis Trial (ARTIST) similar procedural results but a lower angiographic restenosis rate for balloon PCI (51%) than RA (65%, P < 0.05) were reported. RA “supporters” argued against trial design, which forced researchers to use low-pressure balloon inflation after RA. Reasons for RA disappointing results in the ARTIST trial were clearly highlighted in an IVUS subanalysis: volumetric lumen gain was 79 ± 68 mm 3 after balloon PCI performed at 13 ± 4 atm, as compared to 44 ± 26 mm 3 after RA and adjunctive PCI performed at a significantly lower pressure 7 ± 3 atm ( P < 0.0001). After RA and low-pressure balloon PCI, stent lumen was 47% smaller as compared to high-pressure stand-alone balloon PCI: combining neointinal compression and stent expansion, net lumen gain after balloon PCI was 82% higher compared to RA .
| Year | Study design | Main Results | |
|---|---|---|---|
| Mehran ( n = 249) | 2000 | RA vs ELCA | Similar procedural success; RA produced a significantly greater reduction in intimal hyperplasia than ELCA (by IVUS, P < 0.001); similar 6-month TLR. |
| ARTIST ( n = 298) | 2002 | RA with ≤ 6 atm balloon PCI vs balloon PCI | Similar procedural success; 6-month restenosis 65% vs 51% ( P < 0.05) |
| ROSTER ( n = 200) | 2004 | RA with 4-6 atm balloon PCI vs balloon PCI | Similar procedural success, but repeat stenting in 10% vs 31% ( P < 0.001); 9-month TLR in 32% vs 45% ( P < 0.05). |
Later on, in the Randomized trial of Rotational Atherectomy Versus Balloon Angioplasty for Diffuse In-stent Restenosis (ROSTER) trial a similar procedural success rate with a higher need for repeat stenting for balloon PCI (31% vs 10% for RA, P < 0.001) and a higher 9-month TLR for balloon PCI (45% vs 32% for RA, P < 0.05) was documented.
Despite the different strategies for the treatment of ISR, the recurrence rate remained high, and only the adjunctive inhibition of neointimal hyperplasia obtained with intravascular radiation therapy (IVRT)—with both γ and β sources—successfully reduced TLR after intervention for ISR . Among different devices used to obtain lumen gain before irradiating neointima, RA showed again some benefit, and a reduced TLR (11%) as compared to additional stenting (22%, P < 0.05) was documented in a large series of patients who underwent PCI with adjunctive IVRT for ISR .
2.6
Rotational atherectomy in the current era
A dramatic reduction in restenosis has been observed with the development of drug-eluting stents (DES)—coated stents capable of releasing antiproliferative agents into the surrounding tissues . However, initial enthusiasm was dampened with the documentation that delayed endothelialisation of the stent struts , inflammation and hypersensitivity reactions might increase the susceptibility of DES to late thrombosis, as documented in the Basel Stent Kosten-Effektivitäts Trial-LAte Thrombotic Events (BASKET-LATE) . Notably, in calcified lesions several procedural concerns are related to DES deployment: a) vigorous manipulation of DES through the stenosis can result in the disruption of the polymer coating and decrease its effectiveness in preventing restenosis, b) a higher rate of delivery failure has been reported with DES as compared to BMS (5.8% vs 2.4%, P < 0.05) , c) suboptimal deployment of DES with strut malapposition despite the use of high-pressure balloon inflations may further increase the risk of stent thrombosis . These limitations of DES usage might be overcome by previous debulking , and RA “equalized” outcomes in a consecutive series of severely calcified lesions pretreated with debulking as compared to non-calcified lesions where only balloon PCI was performed prior to DES implantation, obtaining similar TLR in both lesion subsets . In single series of patients presenting with significant comorbidities and high-risk features, the combined approach of RA and DES had a favorable effect when dealing with heavily calcified lesions, without major safety concerns .
The recently completed Rotational Atherectomy Prior to Taxus Stent (ROTAXUS) is the only randomized trial that evaluated whether routine debulking was more effective than standard balloon pre-dilation prior to DES deployment in complex calcified lesions: as compared to balloon pre-dilation, RA obtained a higher procedural success (92.5% vs 83.3%, P = 0.03). The higher acute gain obtained with RA (1.57 ± 0.43 vs 1.46 ± 0.46 mm, P = 0.03) was counterbalanced by an augmented late loss (0.44 ± 0.58 vs 0.31 ± 0.52 mm, P = 0.01), failing to document any difference in restenosis rate (11% vs 11%, P = NS). Minor critiques to the trial may partly justify the failure of any benefit documented with RA: severe calcifications were present in less than half of the study population and in the standard balloon arm the rate of B2/C lesions was significantly lower (86%) than in the RA arm (94%, P = 0.03).
In extremely high-risk subjects a strategy of debulking prior to stent deployment may be proven beneficial ; among hemodialysis patients with calcified coronary lesions, RA-assisted DES deployment has been associated with a lower TLR (11.5%) as compared with standard DES implantation (36%, P < 0.05) and a trend towards a reduced subacute thrombosis rate (0% vs 7.1%, P = NS) .
Beyond planned reduction of calcification before stent deployment, RA has been also used as a last resource in cases of stent underexpansion in calcified plaques , where—after failure of high-pressure and cutting balloon inflations—the ablation of the stent-calcium complex may be effectively obtained with rotablation ( Fig. 3 ).
Another “niche” application of RA has been the treatment of chronic total occlusions (CTO), where even with dedicated technology a low procedural success rate is reported . Here, the most common reason for failure is the inability to cross the lesion, followed by the inability to advance or dilate any balloon over the guidewire correctly positioned through the occluded segment. In the pre-stent era the Balloon Angioplasty versus ROtational angioplasty in Chronic Coronary Occlusions (BAROCCO) study documented a similar procedural success among RA and balloon PCI. After successful wire crossing, Tsuchikane et al. assigned 266 patients with CTOs to balloon predilation or debulking with either RA or DCA before stenting in the Debulking Of Chronic coronary Total Occlusions with Rotational or directional atherectomy before Stenting (DOCTORS) study ; the 1-year adverse event rate was lower in the debulking group than in the non-debulking group (27.5% vs 40%; P = 0.033) More recently, RA successfully overcame balloon failure in 43/45 patients (96%) with CTO resistant to recanalization by conventional techniques, allowing effective deployment of DES . Particular care must be devoted to distal embolization, as coronary collateral circulation can be damaged either by direct collateral injury or through the de-recruitment that occurs early after the re-establishment of antegrade flow .
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
