Clinical expert consensus document on rotational atherectomy from the Japanese association of cardiovascular intervention and therapeutics: update 2023

The Task Force on Rotational Atherectomy of the Japanese Association of Cardiovascular Intervention and Therapeutics (CVIT) proposed the expert consensus document to summarize the techniques and evidences regarding rotational atherectomy (RA) in 2020. Because the revascularization strategy to severely calcified lesions is the hottest topic in contemporary percutaneous coronary intervention (PCI), many evidences related to RA have been published since 2020. Latest advancements have been incorporated in this updated expert consensus document.


Introduction
Severe calcification in atherosclerotic plaques has been the most common cause of poor clinical outcomes since the beginning of percutaneous coronary intervention (PCI) [1][2][3]. Rotational atherectomy (RA) has been widely used for severely calcified coronary lesions for more than 20 years to improve clinical outcomes in patients with severely calcified lesions. Recently, North American expert review as well as European expert consensus on RA have been published to provide a clinical standard for RA operators [4,5]. As compared to North America and European countries, RA in Japan has uniquely developed with the aid of greater usage of intravascular imaging devices [6,7]. Since the cost of intravascular ultrasound (IVUS) or optical coherence tomography (OCT) during percutaneous coronary intervention (PCI) has been covered by the government insurance system in Japan, RA operators could easily access to IVUS or OCT. IVUS has been used to understand the guidewire bias and to decide appropriate burr sizes during RA [8], whereas OCT can be used to measure the thickness of calcification during RA [9], which could result in appropriate burr size up [10].

3
Furthermore, the prevalence of PCI with RA has been higher in Japan than in other countries. In fact, the prevalence of PCI with RA was approximately 3.3% in Japan [11], which was similar to those in United Kingdom (3.1%) or France (2.9%) and was higher than those in Italy (1.3%) or Germany (0.8%) [5].
On the other hand, RA had not been allowed to any operators without on-site surgical back-up in Japan until April 2020. However, the government and the Japanese Association of Cardiovascular Intervention and Therapeutics (CVIT) worked together, and released the new facility criteria for RA in April 2020, which allowed operators to perform RA without on-site surgical back-up. The new facility criteria has rapidly increased the number of RA operators in Japan. In fact, more than 300 facilities have completed the RA training program that was organized by CVIT and manufacturer (Boston Scientific Japan). Early experience from a new comer facility was reported [12], suggesting the safety and feasibility of RA in new comer facilities. The Task Force on Rotational Atherectomy of CVIT proposed the expert consensus document on RA to summarize the techniques and evidences regarding RA in 2020 [13]. Because the revascularization strategy to severely calcified lesions is the hottest topic in contemporary PCI, many evidences related to RA have been published since 2020. Latest advancements have been incorporated in this updated expert consensus document.

Aim of RA
Before the stent era, the main purpose of RA was to debulk atherosclerotic plaques including calcification in coronary arteries [14]. However, the incidence of restenosis was considerable following debulking using RA [15]. Emergence of drug-eluting stent (DES) has dramatically changed the indications of PCI, which included diffuse long calcified lesions. The lesion modification, which facilitate the delivery and expansion of DES, would be the most frequent purpose in the contemporary RA, and the long-term outcomes of DES following RA was acceptable [16][17][18][19][20][21][22], except specific lesions such as calcified nodule [23,24]. The mid-term or long-term clinical outcomes following RA might not be satisfactory in specific characteristics such as hemodialysis or malnutrition [25][26][27], partly because the natural prognosis of patients with such characteristics would be worse than without such characteristics. Incomplete stent apposition (ISA) is frequently observed in severely calcified lesions or calcified nodules even after the lesion modification by RA [28,29], while the clinical significance of ISA in severely calcified lesions remains undetermined.
Moreover, RA might prevent the polymer damage, when DES was delivered to the calcified lesions [30]. Although the initial results of debulking using RA was not satisfactory, the debulking might have developed with the aid of imaging devices and drug-coated balloon in Japan [31][32][33][34][35]. Furthermore, the long-term outcomes following RA with plain old balloon angiography was acceptable when the target lesions were limited to large coronary arteries (diameter ≥ 3 mm) [36]. Intravascular lithotripsy (IVL) is a novel technique to treat severely calcified lesions. Although the combination of RA and IVL has not been systematically examined, RA may facilitate the IVL catheter to reach the severely calcified lesions [37,38]. The contemporary indications for RA is summarized in Table 1. In IVUS, calcifications are typically described as high echoic signals with an acoustic shadow behind the signal [39]. Reverberations that are concentric and arctic lines at equal intervals would be observed within the acoustic shadow [39,40]. Because reverberations are closely associated with smooth surface of the calcified lesions [41], reverberations may be observed in the calcified lesions after RA as well as the unmodified calcified lesions with smooth surface. Recently, Jinnouchi, et al. reported the association between slow flow following RA and reverberations [42]. However, the clinical significance of the natural reverberations has not sufficiently investigated. A systematic review and meta-analysis compared the clinical outcomes between planned RA and bailout RA, and reported that planned RA resulted in significantly shorter procedural times, less contrast use, lesser dissection rates and fewer stents used as compared to bailout RA [43].

Junior RA operator
In this document, we defined junior RA operator as RA operator with less than 50 RA experiences. Because the number of RA cases per operator as well as the number of RA cases per year in the facility was inversely associated with adverse events [11,44,45], the number of RA cases per operator would be important to prevent severe complications. Moreover, the good indication for RA is not necessarily low-risk. For example, the ostial right coronary artery (RCA) lesions have been recognized a good indication for RA [46], whereas RA to the ostial RCA is known to be technically difficult [47]. Therefore, the recommendation for junior RA operators may be different from that for senior operators in some sections. is not allowed to use as a support device for patients who undergo elective high-risk PCI in Japan, the Impella is considered to be an option as a support device for elective PCI with RA in USA [49]. If patients with cardiogenic shock already received the Impella or V-A ECMO supports, RA to the severely calcified lesions can be a bailout option from cardiogenic shock. Furthermore, although the beta-blockers are cornerstone for optimal medical therapy, some operators hesitate to use beta-blockers, because of the possible risk of slow flow [50]. Of course, bradycardia (heart rate < 60 bpm) could be a problem during RA, beta-blockers can be continued, because the risk of slow flow was comparable between with and without beta-blockers [51]. Arrhythmia such as bradycardia or atrioventricular block sometimes happens in RA, especially during the treatment of RCA. Temporary pacing is a reliable option to continue procedures during arrhythmia. However, senior RA operators may not use temporary pacing by several reasons: (1) short ablation time may not induce sustained arrhythmia, (2) cough resuscitation may be effective for arrhythmia during RA [52], (3) there is a risk of ventricular perforation induced by temporary pacing catheter [53,54]. Nevertheless, it would be a safe approach for junior RA operators to insert temporary pacing for specific lesions. Brady-arrhythmia can occur during RA to RCA, dominant left circumflex (LCX), left main trunk, and rarely left anterior descending artery (LAD) lesions. In this document, we recommend junior RA operators to consider temporary pacing for RA to RCA, dominant LCX, and left main trunk lesions.
Kusumoto et al. reported a unique case of trans-coronary pacing via RotaWire [55]. In this case report, the cathode of an external pacemaker was attached to the distal external end of the guide wire using a crocodile clip, whereas the anode was attached to the needle which is inserted under the skin of the anaesthetized groin [55]. The possibility of temporary guidewire pacing can be discussed as potential substitutes for temporary pacing catheters. However, since there are several limitations such as coronary spasm or twitching diaphragm in temporary guidewire pacing, we do not recommend temporary guidewire pacing in this document yet.

Guide catheter for RA
Although RA is possible either trans-radial, trans-femoral, or trans-brachial [56][57][58], RA operators should recognize the maximum burr size for each guide catheter size. A 6-Fr guide catheter can accommodate ≤ 1.75 mm burr, where as a 7 Fr guide catheter can accommodate ≤ 2.0 mm burr. When RA operators consider ≥ 2.15-mm burr, a 8 Fr guide catheter is necessary. However, if there is a severe tortuosity in a guide catheter, operators may feel a strong resistance during advancing the burr in the guide catheter, which results in the burr-size down. Moreover, if operators make side-holes in a guide catheter by themselves, such handmade side-holes may prevent the burr from advancing in the guide catheter. For junior RA operators, it is important to check coronary flow more frequently than senior RA operators to notice the occurrence of slow flow or perforation immediately, which would be easier in ≥ 7 Fr guide than in 6 Fr guide because of the large dimeter of the drive shaft sheath (4.3 Fr).
The choice of guide catheter curves varies even among senior RA operators. Although the appropriate back-up support is important for stable procedures, the strongest back-up support, which is sometimes required for PCI to chronic total occlusion, would not be necessary for RA. The coaxial positioning of the guide catheter would be of utmost importance for successful RA. However, the strong back-up support can be a key to success in limited cases [59].

Guidewire for RA
Two types of guidewire for RA have been commercially available: RotaWire floppy (Boston scientific, Marlborough, MA, USA) and RotaWire Extra-support (Boston scientific, Marlborough, MA, USA). Both RotaWires have 0.014-inch/0.36 mm spring tip and 0.009 inch/0.23mm guidewire shaft. However, the length of the tapered segment is considerably different between RotaWire floppy and RotaWire Extra-support. RotaWire floppy has long tapered shaft (13 cm of < 0.0077-inch/0.20 mm shaft) and short spring tip (22 mm), whereas RotaWire Extra-support has short tapered shaft (5 cm of < 0.009inch/0.23 mm shaft) and long spring tip (28 mm) [60]. For successful RA, it is important not to make a bend in RotaWires, because a bend in RotaWires substantially increase the friction force between burr and RotaWire. Therefore, the use of microcatheter is recommended to bring RotaWires to the target. RA operators should use the conventional 0.014-inch guidewire to cross the lesion, and then exchange the conventional guidewire to the RotaWire using microcatheter. Recently, the manufacturer (Boston scientific, Marlborough, MA, USA) refined both RotaWire floppy and RotaWire Extra-support, and introduced both RotaWire drive floppy and RotaWire drive extra-support [61]. The torque response has considerably improved in both RotaWire drive floppy and extra-support. However, we still recommend to use the conventional 0.014-inch guidewire to cross the lesion, and then exchange the conventional guidewire to the RotaWire drive using microcatheter not to make a bend in RotaWires. The guidance for selection of RotaWires is shown in Table 2.

Appropriate burr size
In early experiences with RA, big burrs were used to debulk the calcified plaques. However, a randomized trial comparing small burrs (burr/artery ratio of ≤ 0.7) with large burrs (burr/artery ratio of > 0.7) revealed that small burrs achieved similar immediate lumen enlargement and late target vessel revascularization compared with large burrs, but showed fewer complications [62]. Furthermore, a recent retrospective study reported that bur/artery ratio > 0.61 was associated with worse clinical outcomes [63]. European expert consensus document recommend burr/artery ratio of 0.6 [5], and North America expert consensus document recommend burr/artery ratio of 0.4-0.6 [4]. In this document, we recommend burr/artery ratio of 0.4-0.6 without intravascular imaging devices, and recommend to use intravascular imaging if RA operators aim to achieve burr/artery ratio of ≥ 0.6. Table 2 Guidance for selection of RotaWires

Characteristics or specific situations
RotaWire floppy or extra-support Ability to ablate the severely calcified plaques (ablation efficiency) Extra-support > floppy Ability to straighten the tortuous coronary artery Extra-support > floppy Ability to strengthen the back-up force in the system Extra-support > floppy When pre-intravascular imaging devices cross the lesion and provide sufficient information regarding the guidewire bias Select either Extra-support or floppy according to the information from imaging devices and angiography When operators cannot judge the guidewire bias from angiography and/or intravascular imaging For lesion modification, single burr may be sufficient to facilitate stent delivery and stent expansion. However, second burr is sometimes necessary even for lesion modification. The guidance for second burr is summarized in Table 3.

Burr manipulation and rotation speed
A pecking motion (quick push-forward/pull-back movement of the burr) has been a standard burr manipulation in RA [5]. Although several burr manipulations have been conducted by RA experts, the common part of burr manipulation is to push-forward from the platform and pull-back the burr to the platform. The speed of manipulation varies widely among experts. Some experts prefer very quick, whereas other experts prefer very slow. Either speed is acceptable as long as the following points are considered: (1) Operators should control the burr's motion. If operators feel difficult to control the burr's motion, the speed may be too quick, (2) operators should avoid excessive rotational speed down, and (3) operators should not deactivate the system when the burr is in the middle of stenosis, which may result in the entrapment of the burr. Operators should deactivate the system when the burr was pulled-back to the platform. Furthermore, the ROTAPRO™ system (Boston scientific, Marlborough, MA, USA) has been gradually introduced to new comer facilities as well as existing facilities in Japan. In brief, the ROTAPRO™ system is composed of a digital console and an advancer including buttons to activate devices [61]. The ROTAPRO™ system eliminated the foot pedal.
Duration of individual runs is also important to prevent complications. Manufacturer recommend the duration of individual runs less than 30 s. In general, longer duration would be associated with greater amount of debris. For junior RA operators, short duration (e.g. ≤ l5-20 s) for a single session would be recommended. Furthermore, it is important to check the situations such as ECG and vital signs between the sessions.
Regarding the rotational speed, since manufacturer set the minimum speed as 140,000 rotations per minute (rpm), and the maximum speed as 190,000 rpm [64], this consensus document also recommends to use 140,000-190,000 rpm, and may consider to use > 190,000 rpm when operators feel difficulty to cross the lesion. There was a debate whether low rotational speed can reduce the risk of slow flow. Platelet aggregation was greater in high-speed (180,000 rpm) than in low-speed (140,000 rpm) in an early experiment in vitro [65], which has not been proved in vivo. Recently, a randomized control study comparing low-speed (140,000 rpm) with high-speed (190,000 rpm) revealed that the incidence of slow flow was similar between low-speed and highspeed [66]. Therefore, it is not reasonable to use low-speed for the prevention of slow flow. On the other hand, there were several interesting findings from Japan regarding the additional lumen gain in low-speed RA. Mizutani, et al. reported that the greater debulking area following low-speed (< 150,000 rpm) was confirmed by OCT [67]. Yamamoto, et al. also reported that the greater debulking area following very low speed (110,000 rpm) was confirmed by OCT [68]. However, Kobayashi, et al. reported that there were no additional lumen gain following low speed (120,000 rpm) [69]. Considering the above evidences, low speed (140,000 rpm) within the instructions for use could be an option to acquire additional lumen gain, but very low speed (< 140,000 rpm) should not be used, especially for junior RA operators. The very-high speed (> 190,000 rpm) is sometimes used in Japan [9,70]. A bench test showed that the RotaWire may be spinning under the maximum rotational speed [71], while the RotaWire theoretically would not spin during highspeed mode because of the internal brake and WireClip. The spinning of RotaWire may be associated with the guidewire failure [72,73], which have not been proved in the large registry data.

How to bring the burr to the platform
Before activating the burr, it is a key to success to bring the burr into the platform with keeping the RotaWire stable position. However, several troubles were frequently observed in the above process. The RotaWire could advance too deep or be pulled-back. The RotaWire may make a loop at the outside of the guide catheter, which can result in severe complications [74]. Although the manufacturer recommends to use the dynaglide mode only when operators remove the burr, not a few RA operators use the dynaglide mode when operators bring the burr to the platform. In this document, we would like to show the risk and benefit of both ways (using dynaglide or not) in Table 4. Because both ways have some disadvantage, operators need to understand both ways to avoid possible complications.

Rota cocktail
RA advancer has a saline infusion port. Although the instructions for use does not recommend to use any drugs into the saline bag, various drugs have been used to prevent slow flow. A representative combination of drugs was verapamil 10 mg (5 mg), nitroglycerin 5 mg (2.5 mg), heparin 10,000 unit (5000 unit), and saline 1000 ml (500 ml). Another representative combination of drugs was nicorandil 24 mg (12 mg), nitroglycerin 5 mg (2.5 mg), heparin 10,000 unit (5000 unit), and saline 1000 ml (500 ml). Two randomized studies compared nicorandil based cocktail with verapamil based cocktail, and showed that the incidence of slow flow was significantly lower in the nicorandil based cocktail than in the verapamil based cocktail [75,76]. Preferred cocktail varied widely among RA experts, partly because intra-coronary injection of vasodilators such as nicorandil or nitroprusside was easily performed in the contemporary catheter laboratories. Either combinations of drugs are acceptable, as long as intra-coronary injection of vasodilators are available in a catheter laboratory. If only saline is used for RA, the activated coagulation time should be checked before RA to prevent possible thrombus formation.

Imaging devices in RA
Imaging devices such as IVUS or OCT is useful in RA [77][78][79][80][81][82][83][84]. In this document, we recommend to use IVUS or OCT before, during, and after RA. Since both IVUS and OCT have advantages and disadvantages, each device should be selected according to the purpose of RA in each case. The advantages and disadvantages of IVUS and OCT are summarized in Table 5. Recently, Kawaguchi, et al. reported the impact of the degree of guidewire bias in the vessel's healthy portion on coronary perivascular trauma in RA [85], which suggests the greater risk of vessel perforation when the intravascular imaging catheter is pushing the normal vessel to distort the vessel structure (tenting phenomenon). Moreover, Hashimoto, et al. also reported that there is a higher risk of medial injury due to the RA procedure, especially near the bifurcation of the left anterior descending artery and diagonal branch when the guidewire and IVUS catheter are close to the healthy side of the vessel wall [86]. Although intravascular imaging before RA can provide useful information regarding appropriate burr size or RotaWire, intravascular imaging catheter may not cross the severely calcified lesions. If pre-RA imaging is critical for safe RA (e.g., ostium of left circumflex lesions), the use of guide-extension catheters may be considered to cross the lesion. For imaging device uncrossable lesions, small burrs (1.25 mm or 1.5 mm burrs) would be the choice. Either 1.25 mm or 1.5 mm burrs is to be decided at operator's discretion. Sakakura, et al. compared the complications between 1.25 mm and 1.5 mm burrs for IVUS-uncrossable lesions, and showed the incidence of complications was comparable [87]. Moreover, the risk of complications was greater in imaging device uncrossable lesions than in imaging device crossable lesions [88]. Junior RA operators should be careful about those imaging device uncrossable lesions. Small balloon dilatation before RA can be an option for junior RA operators to prevent the severe complications.
Recently, Maehara and colleagues developed the IVUS calcium score as well as the OCT calcium score [89,90]. The OCT calcium score was composed of maximum calcium angle (> 180º = 2 points), maximum calcium thickness (> 0.5 mm = 1 point), and calcium length (> 5 mm = 1 point) [89]. The OCT calcium score was closely associated with stent underexpansion [89], which suggests that the lesions with high OCT calcium score would require RA to achieve optimal stent expansion. The IVUS calcium score was composed of length of superficial calcification > 270º (≥ 5 mm = 1 point), presence of 360º superficial calcification (yes = 1 point), presence of calcified nodule (yes = 1 point), and vessel diameter (< 3.5 mm = 1 point) [90]. They advocated that operators should consider the use of atherectomy devices to avoid stent underexpansion when the target lesion's IVUS calcium score was ≥ 2 points [90]. Although both IVUS and OCT calcium scores have a potential impact on the indication for RA, intravascular imaging catheters may not cross the severely calcified lesions before RA. These calcium scores may be more applicable to decide the indication for the burr size up or other aggressive lesion modifications after the small burr cross the lesion rather than to decide the indication for RA.

Complications: slow flow
Slow flow is the most frequently observed complication following RA. The incidence of slow flow was approximately 5-20% [66,76,92], and varied widely among literatures, partly because the timing of judgement (just after RA or final shot) and the definition of slow flow were different among literatures. Lesion length and burr-to-artery ratio were reported as the determinants of slow flow [66,93]. Furthermore, IVUS findings such as longer lesion length, the maximum number of reverberations, and the greater arc of calcification at MLA may predict slow flow after RA [42]. For the prevention of slow flow, appropriate burr size, short ablation time, and gentle manipulation avoiding excessive speed down would be important to minimize the amount of debris caused by RA. A retrospective study revealed that short single session (≤ 15 s) was inversely associated with slow flow after RA [93], which suggests that short single session can reduce the incidence of slow flow after RA. A randomized control study to compare the incidence of slow flow following RA between the short single session strategy and the long single session strategy is currently ongoing [UMIN000047231]. Although there were no literatures, some experts prefer to flash saline for the prevention of slow flow during RA. Moreover, it is important to keep sufficient systolic blood pressure ≥ 120 mmHg (at least 100 mmHg). If patient's cardiac function is normal, sufficient hydration is also important. If a patient shows low blood pressure under poor cardiac function, IABP may be considered. Noradrenaline diluted in saline is frequently used to keep blood pressure. If operators took a venous sheath from femoral vein as a rescue sheath, such sheath would be helpful to inject the drug immediately. Unlike slow flow during primary PCI, slow flow during RA is gradually developed (i.e. TIMI-3 flow to TIMI-2 flow, then TIMI-2 flow to TIMI-1 flow, etc.) as long as operators do not ablate lipid-rich plaques. Therefore, it is important to watch ST-elevations in ECG, which usually antecedent slow flow. If the TIMI-2 slow flow occurs, RA should be stopped temporarily until the TIMI-3 flow was restored. Most transient TIMI-2 slow flow would not result in periprocedural myocardial infarction if treated immediately [94]. If blood pressure fall following slow flow, noradrenaline diluted in saline is injected to restore blood pressure. If noradrenaline did not work, the prompt insertion of IABP would be a next option. Intra-coronary vasodilators such as nitroprusside, nicorandil, and nitroglycerine are used to treat slow flow. Although there were no literatures comparing the efficacy among such vasodilators, nitroprusside may be the most potent vasodilator for slow flow [95].
Although nicorandil may be more effective than nitroglycerine [96], the rapid injection of nicorandil may provoke fatal arrhythmia or even cardiac arrest [97]. The use of microcatheters or double lumen catheters would be considered to minimize the risk of vasodilator-induced hypotension. Appropriate timing of intra-coronary vasodilators would be important to treat slow flow. Operators should check ECG or vital signs between sessions, and check the flow when the change of ECG was observed. In case of severe slow flow (TIMI-0), the use of thrombectomy catheter can be considered before the injection of intra-coronary vasodilators. The prevention and bailout for slow flow are summarized in Table 6.

Complications: perforation/rupture
Coronary perforation due to the burr is the most serious complication in RA, and the incidence of perforation in RA is approximately 1-2% [98][99][100][101]. If coronary perforation occurs during RA, there would be a considerable risk of in-hospital death, especially the lesions involving left main coronary artery [102]. The risk of perforation highly depends on the lesion characteristics such as vessel tortuosity or eccentricity of calcification. Since the shape of each burr is ellipsoid [103], the RA burr cannot follow the sharply angulated vessel, which results in the greater risk of perforation. The risk of perforation is generally considered to be greater in an eccentric calcification such as calcified nodules than in a concentric calcification such as napkin-ring calcification. Thus, operators need to be careful for RA to an eccentric calcification. The selection of appropriate burr size and RotaWire should be important to prevent perforation, and the use of intravascular imaging devices would help operators to select appropriate burr size and RotaWire. If intravascular imaging shows the finding that the guide wire is pushing the normal vessel to distort the vessel structure, the risk of vessel perforation/injury would be greater following RA [85]. If intravascular imaging devices could not cross the lesion, small burrs (1.25 mm or 1.5 mm) should be the choice, especially for junior RA operators. In general, a RotaWire floppy would follow the vessel without distorting the vessel configuration, whereas a RotaWire Extra-support would follow the vessel with distorting the vessel configuration. Therefore, the route of the burr would be considerably different between RotaWire floppy and RotaWire Extrasupport, which suggests the importance of the choice of RotaWires for the prevention of perforation. If the guidewire bias was difficult to anticipate by intravascular imaging or angiography, a RotaWire floppy would be the choice. The bailout of perforation caused by RA is basically similar to that caused by PCI without RA except the fact that operators have to remove the RA system with keeping the RotaWire. If operators lost the RotaWire following massive perforation, there would be no guarantee of recrossing the guidewire. A Kusabi trapping balloon (Kaneka, Osaka, Japan) can be used for the retrieval of ≤ 1.5 mm burrs in ≥ 7 Fr systems [104]. A balloon catheter or a perfusion balloon catheter should be promptly delivered to the lesion to seal the blood flow toward pericardial space. Since the pericardiocentesis is probably necessary in perforation following RA, the preparation for pericardiocentesis is important in catheter laboratories using RA. Multiple covered stents may be necessary to seal the blood flow [105,106]. The use of guide extension catheter or the use of double guide catheters may be considered to facilitate the delivery of covered stents [107,108], because there would be a risk of dislodgement of covered stents in severely calcified lesions. In the use of double guide catheters, a balloon via the first guide catheter can be used to seal bleeding during the preparation of covered stent via the second guide catheter, and the same balloon can be used as the distal anchor balloon to facilitate the covered stent delivery. Moreover, operators should contact to cardiovascular surgeons immediately just in case of unsuccessful percutaneous bailout. The prevention and bailout for coronary perforation are summarized in Table 7.

Complications: burr entrapment
Burr entrapment is a unique complication in RA, and the incidence of burr entrapment is not derived from multicenter registries, but is available from single center studies ranging 0.4-0.8% [109,110]. Burr entrapment can occur from several mechanisms. One is called as "Kokesi phenomenon" that the burr was trapped in the distal portion of the proximal narrowing [110]. The mechanism of Kokesi phenomenon is considered to be that a friction heat enlarges the orifice and the coefficiency of friction in motion is smaller than that of friction at rest [110]. This type of burr entrapment may occur following forceful manipulation with small burrs. Another mechanism is the burr entrapment related to the vessel angulation. Since the shape of the burr is ellipsoid and the diamond coating is not available at the tail of the burr, the burr can be trapped by non-massive calcification at the site of angulation. To prevent burr entrapment, operators need to be careful about rotational speed reduction, sound of ablation, and resistance during the burr manipulation. If operators encounter the burr entrapment, it is important to assess the situation calmly. The presence of antegrade flow, ST-segment elevations in ECG, and patient's chest pain should be evaluated. If there is no antegrade flow beyond the entrapped burr, the percutaneous bailout would be very difficult, and be limited to experienced senior RA operators.
In the meantime, operators should contact cardiovascular surgeons to discuss the surgical bailout. If the antegrade flow is present without ST-segment elevations, operators would have a time to consider the percutaneous bailout techniques. Although there have been several percutaneous bailout techniques in literatures [111][112][113][114][115][116][117][118], the main difference among various techniques was whether to use additional guide catheter (double guide catheters) or not (single guide catheter). If operators selected the double guide systems, operators would insert the conventional guidewire from the second guide catheter, and then would try to dilate the proximal part of the entrapped burr using a balloon [112]. If operators selected the single guide system, the next step would depend on the guide catheter size (≥ 8 Fr or ≤ 7 Fr). If operators used a ≥ 8 Fr guide system, operators would insert the conventional guidewire, and then would try to dilate the proximal part using a balloon. However, if operators used a ≤ 7 Fr guide system, operators need to cut and pull the drive shaft sheath (Fig. 1) [111], because a ≤ 7 Fr guide catheter cannot accommodate the drive shaft sheath, guidewire, and balloon catheter together. After pulled out the drive shaft sheath, operators can try to dilate the proximal part using a balloon. Once operators pulled out the drive shaft sheath, operators can use inner catheters such as guide extension catheters [113,116,117]. Of course, operators should recognize the possibility of unsuccessful percutaneous bailout, and need to prepare the massive perforation or severe dissection following the burr retrieval [119]. The prevention and bailout for burr entrapment are summarized in Table 8.

Complications: transection of the RotaWire
The transection of the RotaWire is a rare complication in RA, and the incidence of transection of the RotaWire is not derived from multicenter registries, but may be approximately 0.4-1% [120,121]. There are two types of the transection of the RotaWire: One is the transection at the radiopaque part of the RotaWire, and the other is the transection at the radiolucent part of the RotaWire. The transection of the radiopaque part is easy to notice. If there were the transection of the radiopaque part of the RotaWire, operators should exchange the broken RotaWire to the new one. Since the retrieval of a transected fragment of the RotaWire would be similar to that of the conventional guidewire, operators might try retrieval procedures such as twin guidewire method [120]. If the transected fragment of the RotaWire located at the far distal segment of the treated vessel, operators might leave it at the distal segment rather than retrieval. On the other hand, the transection of the radiolucent part of the RotaWire is very difficult to notice. If operators could not notice the transection of the radiolucent part of the RotaWire, operators would have a vessel perforation [121][122][123]. In fact, Wang et al. reported that the Rotawire damage with subsequent transection was the cause in 18.2% of cases with coronary perforations [101]. Moreover, there would be a greater risk in retrieval of a transected fragment of the RotaWire, if the transection occurred at the radiolucent part. Because the proximal part of the transected fragment of the RotaWire would be sharp, the invisible sharp fragment of the RotaWire may damage to the proximal vessel wall, even to the aortic wall during the retrieval. IVUS may be helpful to identify the invisible fragment.
Since the incidence of transection of the RotaWire is low, the causes of transection have not fully understood. Because the metallic fatigue would be the possible cause of transection, it may be important to avoid the continuous contact between the burr and the specific part of the RotaWire. If the RotaWire kinked at the outside of the guide catheter, the transection of RotaWire would happen when the burr advanced over the kinked RotaWire [74]. Furthermore, the WireClip Torquer (Boston Scientific, Marlborough, MA, USA) should be equipped with the RotaWire during the dyna-glide mode as well as the high-speed mode to avoid the spinning of the RotaWire, which could be a cause of transection of the RotaWire.

Complications: disconnection of the burr
Disconnection of the burr is very rare complications, and there have been no literatures mentioning the incidence of disconnection of the burr. The exact reasons or mechanisms of disconnection of the burr has not been specified. Even senior RA operators with abundant experiences may not or may have a few cases with this complication. Theoretically, the disconnection of the burr could happen when operators activate the burr after the burr was entrapped, or when operators advanced the burr in spite of strong resistance within the stented segment or angulated segment. Operators may notice the disconnection of the burr by the loss of coordination between the burr motion and the nob motion. The percutaneous bailout may be possible if the RotaWire is not transected, because the 0.014-inch spring tip of the RotaWire may work as the anchor. Imamura S, et al. reported a case of the disconnection of the burr, which was successfully treated percutaneously [124]. In their case report, since the simple pull-back using a guide extension catheter did not work, they sandwiched the disconnected burr between the inflated balloon and the guide extension catheter, and then pulled back together [124]. However, surgical bailout should be considered to this rare complication [125].

Specific lesions: RCA ostial lesions
Since clinical outcomes of RCA ostial lesions have been unsatisfactory for decades [126], RA has been considered to be a good indication for RCA ostial lesions with severe calcification [46]. However, a good indication does not necessarily mean an appropriate lesion for junior RA operators. Sakakura, et al. reported that the excessive speed reduction during RA was significantly associated with RCA ostial lesions [47], which suggests the difficulty of RA for RCA The selection of the RotaWire for RCA ostial lesions is also important. The RotaWire Extra-Support may be preferable when operators intentionally remove the guide catheter from the RCA ostium, whereas the RotaWire Extra-Support may be dangerous when operators cannot take a coaxial position. IVUS may help to estimate the risk of RA, especially when operators could not take a coaxial position. Although an IVUS catheter may not cross the lesion before RA, an IVUS catheter would cross the lesion after the crossing of small burrs. IVUS should be tried before using the big burrs for RCA ostial lesions. Furthermore, there may be an additional risk of cerebral infarction following RA for RCA ostial lesions. It may be important for the prevention of cerebral infarction to use small burrs for an initial attempt to minimize the size of debris. Figure 2 summarized the why RA to ostial RCA is difficult. RA for RCA ostial lesions is not recommended for junior RA operators without senior RA operator's back-up.

Specific lesions: LCX ostial lesions with substantial bending
Severely calcified left circumflex (LCX) ostial lesions with substantial bending may be the highest risk lesion in RA. The key to success would be the interpretation of pre-procedural intravascular imaging. If the pre-procedural imaging device could cross the lesion and provide sufficient information including the guidewire bias, operators could select appropriate RotaWires and burrs. However, if the pre-procedural imaging device could not cross, operators need to select RotaWires and burrs from the only angiographical information. In general, atherosclerotic plaques are observed in the lateral wall, whereas atherosclerotic plaques are spared in the flow divider regions (carina) [127]. If severe eccentric calcification was observed in the lateral wall of the LCX ostium, there would be a risk of perforation in the carina side of the LCX ostium due to the jumping of the burr. Operators would try to ablate the lateral wall of the LCX ostium to avoid perforation in the carina side. However, if operators ablated the lateral wall too much, there would another risk of perforation in the lateral wall side due to the deep cut. Two types of perforation in LCX ostial lesions with substantial bending are illustrated in Fig. 3. Operators need to select appropriate burr size, RotaWires, and burr motion to prevent the above two types of perforation. In angiography, it would be important to evaluate the actual contact point between the calcification and the RotaWire using multiple projections. Although low-dose radiation angiography is important for patient's safety, excessive low-dose radiation angiography may sacrifice the visibility of RotaWires or calcification. Since the visibility of RotaWires or calcification is the cornerstone for high-risk RA, operators should set the radiation dose not to prevent the visibility of RotaWires or calcification. Moreover, if there would be a perforation in the LCX ostium, the percutaneous bailout would be difficult, because the implantation of covered stents may occlude the left anterior descending artery (LAD). Therefore, the indication as well as the strategy of RA to the LCX ostium should be carefully discussed. RA for LCX ostial lesions with substantial bending is not recommended for junior RA operators without senior RA operator's back-up.

Specific lesions: unprotected left main lesions
RA to the unprotected left main lesions requires special attention, because slow flow may cause hemodynamic collapse. However, Fuku et al. compared clinical outcomes of left main PCI between with RA (n = 108) and without RA (n = 1091), and showed the similar rate of complications in left main PCI between with RA and without RA [128], suggesting the acceptable safety in PCI to left main lesions using RA. In clinical practice, it would be important to start with small burrs (≤ 1.5 mm burr) to prevent fatal slow flow. However, the ≤ 1.5 mm burr may be too small as the final burr size considering the vessel diameter of left main lesions. It would be important to use IVUS/OCT to select an initial burr size and final burr size. If the severe calcified plaques exist at the ostium of left main, the procedure would be more complex as compared to calcified plaques at the middle of left main. In that situation, the stabilization of guide catheter at co-axial position within aorta would be the key to success like the ostium of RCA lesions. Furthermore, the use of IABP and/or temporary pacing should be considered. RA for left main lesions is not recommended for junior RA operators without senior RA operator's back-up.
If we do not limit the topic to the left main lesions, the subgroup analysis of the PREPARE-CALC study revealed that the incidence of side branch compromise was less Fig. 3 Schema of two types of perforation in LCX ostial lesions with substantial bending frequently observed in lesions treated with RA than in lesions treated without RA in bifurcation lesions [129]. Mizuno, et al. also reported that the incidence of side branch compromise was less frequently observed in calcified bifurcation lesions with RA than in those without RA, even if RA was performed to the main vessel only [130]. On the other hand, Chen, et al. reported the higher risk of side branch perforation, when RA was performed to both main vessel and side branch as compared to RA to the main vessel only [131]. Therefore, RA to both main vessel and side branch should be reserved for senior RA operators.

Specific lesions: stent ablation
Stent ablation is applied to the restenotic lesions caused by under-expanded stents. Okamura, et al. described a case of stent ablation to treat restenosis lesions due to crushing of a sirolimus-eluting stent [132]. In their report, they performed an experimental study and found that the size of the metallic particles generated during RA of stent struts was 5.6 ± 3.6 μm, which suggested the safe size in a human body [132]. Tips of stent ablation are (1) to select the appropriate burr size, (2) to ablate only stent struts first, (3) and then ablate the calcification behind stent struts using the second burr (size up). Since the risk of burr entrapment is greater during stent ablation, gentle manipulation with gradual size-up would be important. Although stent ablation might be a single solution for the restenostic lesions caused by under-expanded stents, the long-term outcomes was not satisfactory [133][134][135][136]. Coronary lithoplasty may be another option for stent underexpansion due to severe calcification [137], while sufficient follow-up data is not available yet. Stent ablation is not recommended for junior RA operators. Furthermore, on-site surgical back-up may be need because of greater risk of burr entrapment, when operators perform stent ablation using big burrs. The burr positioned just before the calcified lesion. E The burr ablated the proximal segment of the calcified lesion. F Balloon dilatation was performed for the rest of the calcified lesion. This figure was reproduced with the permission from Sakakura, et al. [141] Specific lesions: diffuse long lesions RA was frequently performed for diffuse long lesions, and the long-term outcomes of long lesions treated by RA were comparable to those of short lesions treated by RA [138]. However, RA for diffuse long lesions may be difficult for junior RA operators, especially when there is an angle in the middle of diffuse long lesions. The presence of an acute angulation on the lesion is reported to be associated with procedural failure of RA [139]. Sakakura, et al. reported the utility of halfway RA especially for lesions with an angle [140,141]. In brief, the operator does not advance the burr beyond the angle within the lesion to avoid burr entrapment or vessel perforation, and balloon dilatation is performed beyond the angle after RA (Fig. 4). Halfway RA may be useful especially for junior RA operators. If halfway RA did not work (e.g., balloon could not dilate the distal calcified lesion), switch from halfway RA to conventional RA should be considered. Although the other option could be to use the guide extension catheter, RA beyond the guide extension catheter is not straightforward. Another important aspect in RA for diffuse long lesions is to check the speed down of rotational speed. Although the excessive speed down during RA should be avoided, absence of reasonable speed down may mean that the burr does not contact to calcification adequately. If an operator pushes the burr too much in the absence of speed down, there would be substantial risk of burr entrapment or vessel perforation. Therefore, an operator may change the RotaWires to facilitate the contact between the burr and the calcified plaques, switch to balloon dilatation (resultantly halfway RA), or rarely burr size-up to increase the contact area in the absence of reasonable speed down [103,142].
Operators should also take care of ischemia during RA to the diffuse long lesions, because the risk of slow flow is greater in the diffuse long lesions than the short lesions [66]. If operators noticed the signs of ischemia in the middle of diffuse long lesions, size down of the burr would be the reasonable choice. Another option would be to take a short break with removing the burr from the coronary artery to stabilize coronary flow and vital signs. If coronary flow is restored and vital signs are stabilized, operators can safely resume procedures. This document suggests an algorithm for disuse long lesions, especially for junior RA operators (Fig. 5).
The maximum working range of the burr is approximately 70 mm. Operators usually set the nob at 1-2 cm apart from the end, and set the platform at 1-2 cm proximal from the lesion. Thus, the actual working range of the burr is approximately 30-50 mm. If the length of the target lesion was beyond 30-50 mm, operators need to move the platform during the procedure. Although there were few literatures regarding how to move the platform during the procedure, there were several ways among RA experts. The most important point is to ablate the proximal part of the diffuse long lesion sufficiently before moving the platform. Only 1 pass would not be enough. Several passes would be necessary until no additional speed down was observed. Some experts prefer to size up the burr and ablate the proximal part of the diffuse long lesion using big burrs to make a stable platform. After operators ablated the proximal part of the diffuse long lesion sufficiently, operators could move the platform more distally. Some experts prefer to use dynaglide mode to move the platform. Sliding sheath technique, in which operators park the burr to the distal end and then bring the advancer to follow the burr, may be used by some experts. Since the outer diameter of the drive shaft sheath is 4.3 Fr (1.43 mm), the ablation by the burr-1.25 mm may not be enough to move the platform.

Specific lesions: ablation to an entrapped guidewire
An entrapment of coronary guidewires is sometimes observed in complex PCI. Forced pull-back of the entrapped guidewire may damage coronary artery wall or provoke the fracture of guidewire. RA can be an option to cut the entrapped guidewires [143][144][145]. In the comparison of forced pull-back by snare, the coronary artery is not subjected to excessive force if the contact point between the burr and the fractured guidewire is controlled under IVUS guidance [144]. However, since there is a potential risk that the RA burr may be entrapped in the spring wire, ablation to an entrapped guidewire is not recommended for junior RA operators.

Conclusions
In this document, we provided the Japanese style RA, which uses intravascular imaging devices to achieve the maximum efficacy without sacrificing the safety. Our recommendation focused on mainly junior RA operators, but would be helpful for senior RA operators for their more advanced procedures.
Funding None.
Data availability This consensus documet does not include original data.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.