Procedural safety of rotational atherectomy and modified balloon angioplasty: insights from a German national registry

Relevant coronary artery calcification (CAC) is present in 20–30% of patients with coronary artery disease (CAD) [1‐3], [4] and is associated with unfavorable outcomes after percutaneous coronary interventions (PCI) [2, 3, 5‐7]. Clinical management of severely calcified coronary stenoses is challenging and requires the careful selection of dedicated interventional devices based on the anatomy of the target lesion and its underlying calcium morphology [8]. The EAPCI clinical consensus statement in collaboration with the EURO4C-PCR group recommends a step-wise management for patients with heavily calcified coronary stenosis, starting with a thorough imaging assessment of calcified lesions and the corresponding selection of appropriate interventional tools for lesion preparation [9].

Within the last decade, modified balloon-based as well as atherectomy-based-techniques have been increasingly recognized as valuable tools for preparation of calcified coronary lesions during PCI. Modified balloons (MB) allow for a more effective dilatation through focal concentrations of dilating force on cutting or scoring elements of the balloon surface. The cutting balloon (Wolverine Cutting Balloon, Boston Scientific) is a non-compliant balloon with three or four microsurgical blades (atherotomes) mounted longitudinally on its outer surface. Scoring balloons are semi-compliant (AngioSculpt, Philips; NSE Alpha, Braun) or non-compliant (ScoreFlex NC, OrbusNeich) balloons surrounded by external nitinol spiral scoring wires on their surface. Rotational atherectomy (RA, Rotablator System, Boston Scientific) employs an elliptically shaped metallic burr with a diamond-coated distal tip rotating with a recommended rotablation speed between 135 000 and 180 000 rpm that preferentially ablates inelastic tissue, such as calcified plaque material [10].

If moderate to severe calcification is identified in a target lesion based on angiography or coronary computed tomography angiography, according to the current consensus direct stenting should be avoided and crossability of the lesion should be assessed by a small balloon [9]. While the use of cutting/scoring balloons is reserved for balloon-crossable lesions, RA can be employed to lesions, which are accessible for a rotawire (a specialized 0.009 inch guidewire) or microcatheter [4, 9]. Both are recommended options for lesions with superficial calcium [9].

Studies focusing on procedural success demonstrated that cutting balloon angioplasty allows for achieving significantly larger acute lumen gain at lower inflating pressure, especially in aorto-ostial lesions and lesions with evidence of dissections in comparison to conventional balloon pre-dilatation [11‐15]. Similarly, pretreatment with a scoring balloon has been demonstrated to improve final stent expansion compared with conventional balloon angioplasty [16, 17]. Conversely, RA prior to drug eluting stent (DES) implantation can improve the immediate procedural success rate compared to direct stenting or pre-dilatation with modifying balloon angioplasty [18‐21].

However, real-world data about the in-hospital safety of RA or MB techniques are scarce. In this study we therefore analyzed patients’ characteristics and procedural safety of cutting or scoring balloon angioplasty and RA from a German nationwide registry. In-hospital safety parameters for both procedures were compared by a propensity score approach. In addition, we evaluated whether outcomes were dependent on procedure numbers per year performed by a single center.

In our study we evaluated in-hospital outcomes of 32,470 patients undergoing RA or MB angioplasty in the German nationwide cohort from 2017 to 2020. Overall, patients treated by RA had a higher prevalence of risk factors associated with CAC such as male sex, more advanced age, chronic renal disease, and diabetes which can be expected since RA is reported to be used and recommended in lesions with a higher severity of calcification compared to MB angioplasty [25].

While MB angioplasty was performed more frequently than RA in general, annual procedure numbers increased for both interventions in our 4-year observational period. Similarly, a previous report of the National Cardiovascular Data CathPCI Registry analyzing the utilization of coronary atherectomy from 2009 to 2016 found that the numbers of coronary atherectomy has increased over time [26]. These developments may be driven by recent technical optimizations of the RA system [8, 27], the increasing prevalence of coronary calcifications [28] and the increasing confidence of interventionalists in attempting more challenging coronary interventions with the advent of new and improvement of established interventional tools [9].

With regards of in-hospital outcomes of RA and MB angioplasty, we found that the risk of pericardial effusion as well as the adjusted risk of necessity for pericardial puncture was significantly higher in patients undergoing RA. Further, transfusion of more than 5 RBC concentrates was also observed significantly more often in patients receiving RA, suggestive of increased frequency of relevant blood loss. These findings are somewhat expected as RA in general represents the more invasive technique and its use particularly in severely tortuous vessels goes alongside with the risk of coronary perforation [29]. Historically, RA treatment during PCI has been found to be associated with an increased risk of complications, with reported procedural complication rates up to 9.7% [30]. Notably these observations were confounded by a higher preprocedural risk profile of patients treated with RA. Also, increasing operator experience and further technical optimizations such as utilizing smaller burr sizes with a burr/artery ratio of 0.5 and current recommendations to employ shorter burring episodes likely contributes to lower peri-procedural complications in contemporary RA inteventions [29, 31, 32]. Evidence from recent randomized clinical trials emphasized that the use of RA, compared with cutting/scoring balloon-based strategies, is superior in terms of procedural success and not associated with increased intraprocedural or periprocedural complication rates [18, 21]. Notably, in PREPARE-CALC, pericardial effusions were numerically more frequent in patients after RA although these findings did not reach statistical significance like in our real-world study [21]. In line with recent trials, in-hospital adverse cardiovascular events in our study did no differ significantly for MB angioplasty and RA [18, 21].

The superior success rate of RA lesion preparation may contribute to the here observed higher number of implanted stents.

Notably, patients treated with RA spent significantly more days in hospital compared to cutting/scoring balloon angioplasty. While this to some extent may reflect pericardial complications and their management, this may also be explained by a greater disease burden in the overall sicker and older study population of the RA group. Moreover, due to its high technical success rate, RA is frequently used as a bail-out tool in complex coronary diseases, which make up to 20% of calcified coronary lesions in RCTs [18]. Even though recent evidence suggests that unplanned RA is generally associated with favorable outcome when compared to planned RA [33], interventions that involve crossing-over to RA to enable procedural success may require longer procedural time and more frequently demand management of periprocedural complications, which may lead to an increase in hospitalization durations [34, 35]. A total of 1179 cases employed MBs and RA during one hospital stay and thus likely represent such bail-out scenarios.

Finally, we aimed to identify patient characteristic risk drivers and center-volume-dependent outcomes of RA. In our analysis, the risk for the occurrence of pericardial effusion, pericardial drainage and bleeding events was significantly higher in cardiac patients with NYHA III/IV staged heart failure and atrial fibrillation. These findings are in line with others in the literature. Patients with atrial fibrillation undergoing percutaneous coronary intervention are typically at a higher risk of bleeding events due to the need for oral anticoagulation on top of antiplatelet therapy [36]. Further, heart failure has been identified as one of the major risk drivers of bleeding in previous studies [34‐36]. While pericardial effusion itself is a common baseline finding in patients with congestive heart failure [37], impaired LVEF has also been associated with unfavorable in-hospital outcomes after PCI by several studies [38, 39]. Interestingly, we could observe an inverse association of annual procedure numbers with the length of hospitalization and the occurrence of acute cerebrovascular events. In a nationwide retrospective cohort study with 9970 patients from Japan a higher hospital volume of ≥ 31 RAs/year was significantly associated with lower complication rates of RA [40]. In line, recent registry data indicated a higher volume of coronary atherectomy procedures was associated with a lower risk of major adverse events but associated with a small increase in risk of coronary perforation [26]. In our study, we did not find a significant influence of the center volume on the risk of pericardial effusions or the necessity for pericardial drainage indicating that this risk may be attributed to the RA technique and the complexity of CAC itself rather than being dependent on interventionalists experience.

Both RA and MBs are commonly and increasingly used procedures for calcified lesions in Germany. Patients receiving RA were older, had more comorbidities including more severe coronary artery disease compared to patients treated with MBs. Procedural safety regarding bleeding and the occurrence of pericardial effusion is superior for MBs. In centers with high annual rotational atherectomv procedure numbers the risk of acute cerebrovascular events is lower and the length of hospital stay is shorter.