Enhanced Efficacy of Rotational Atherectomy for Calcified Nodules With Contralateral Calcification: Insights From a Multicenter Intravascular Ultrasound Imaging Study

Abstract

BACKGROUND:

Calcified nodules (CNs) represent a high-risk coronary lesion phenotype associated with target lesion revascularization (TLR). Although rotational atherectomy (RA) is an established treatment for calcified lesions, its benefit for CNs remains unclear. This study aimed to evaluate the impact of RA on TLR and to identify specific morphological features on intravascular ultrasound that may influence its therapeutic effect for CNs.

METHODS:

In a substudy of the U-SCAN registry (Coronary Intravascular Ultrasound for Calcified Nodule), 348 patients with CNs identified by intravascular ultrasound who underwent percutaneous coronary intervention were analyzed. We excluded patients with in-stent restenosis, use of alternative debulking devices, failed device passage without RA, and poor image quality. The final analysis included 209 patients, stratified by RA use. Multivariable Cox proportional hazards models were used to identify predictors of TLR and assess treatment interactions across subgroups.

RESULTS:

Among 209 patients, 79 patients (37.8%) underwent RA. During a median follow-up of 2.1 years (interquartile range, 0.4–4.9), TLR was required in 20 of 79 patients (25.3%) in the RA group and 41 of 130 patients (31.5%) in the non-RA group. After adjustment, RA independently predicted reduced TLR (hazard ratio, 0.34 [95% CI, 0.19–0.62], P<0.001). In addition, intravascular ultrasound–derived calcification features, including greater lumen area stenosis, longer CN length, smaller final minimum lumen area, and adjacent circumferential calcification, were significantly associated with TLR. Notably, the benefit of RA on TLR was pronounced in patients with contralateral calcification (8.6% versus 51.6%, P<0.001). In contrast, without this feature, the TLR rate was higher in the RA group (38.6% versus 25.3%, P=0.11), resulting in a statistically significant interaction (P_interaction<0.001).

CONCLUSIONS:

In patients with CNs, RA was associated with a reduced long-term risk of TLR. The presence of contralateral calcification identifies a subgroup deriving substantial benefit, supporting a more selective, morphology-guided approach to treatment.

REGISTRATION:

URL: https://jrct.mhlw.go.jp/; Unique identifier: jRCT1050240037.

WHAT IS KNOWN

• Calcified nodules (CNs) are a high-risk coronary lesion subset associated with poor long-term outcomes after percutaneous coronary intervention.

• Although rotational atherectomy is an established strategy for severely calcified lesions, its efficacy specifically for CNs has been debated.

WHAT THE STUDY ADDS

• This study identifies several intravascular ultrasound–derived features of extensive calcification—including greater lumen area stenosis, longer CN length, and adjacent circumferential calcification—as strong, independent predictors of an increased risk for target lesion revascularization.

• The therapeutic benefit of rotational atherectomy in reducing long-term target lesion revascularization is not uniform across all CNs.

• The benefit of rotational atherectomy is primarily confined to a specific anatomic subgroup: CNs with contralateral calcification, which can be identified by intravascular imaging.

Calcified nodule (CN) represents a particularly challenging phenotype of coronary lesions, which is associated with an increased risk of repeat revascularization despite the use of drug-eluting stent (DES).^1–3 CNs are known to cause target lesion revascularization (TLR) through a distinct mechanism—early reprotrusion of calcified fragments through stent struts,⁴ as well as malapposition and stent underexpansion.⁵ This refractory feature of CN after percutaneous coronary intervention (PCI) suggests the need for an additional therapeutic approach mitigating subsequent risks of repeat revascularization. Although rotational atherectomy (RA) is a primary tool for plaque modification and procedural success in calcified lesions,⁶ its effectiveness for CNs remains a subject of debate. A previous retrospective analysis reported that not using RA was an independent predictor of restenosis, suggesting a potential benefit.⁷ Conversely, a propensity score–matched study directly comparing PCI with and without RA for CNs found no significant difference in clinical outcomes, questioning the routine use of this modality.⁸ This discrepancy may be attributable not only to the morphological diversity of the CN itself, but more importantly, to the heterogeneity of surrounding calcification patterns. This heterogeneity suggests that the benefit of RA is not uniform and may be confined to specific morphological subtypes.

We therefore hypothesized that the efficacy of RA in reducing TLR is critically dependent on lesion morphology, a factor not fully explored in previous reports. Using data from the large, multicenter U-SCAN registry (Coronary Intravascular Ultrasound for Calcified Nodule),⁹ this study aimed to evaluate this hypothesis by assessing the overall impact of RA on long-term TLR in patients with CNs and identifying the intravascular ultrasound (IVUS)–derived features associated with its effectiveness for preventing TLR.

Methods

Data Availability Statement

The data supporting the findings of this study are available from the corresponding author, on reasonable request, due to ethical restrictions related to patient privacy.

Study Population

The present study was conducted using data from the U-SCAN registry, a multicenter observational cohort established at 3 cardiovascular centers in Japan (Methods S1 in the Supplemental Material). The design of this registry has been described in detail previously.⁹ Briefly, between April 1, 2008, and December 31, 2024, the registry consecutively enrolled patients with coronary artery disease who underwent PCI for intravascular ultrasound (IVUS)–derived CN. The following lesions were excluded from the present analysis: (1) in-stent restenosis; (2) cases in which orbital atherectomy, excimer laser coronary atherectomy, or directional coronary atherectomy was used during PCI; (3) lesions in which an IVUS catheter could not be advanced before performing RA; and (4) poor image quality. All eligible patients were categorized into 2 groups based on whether RA was used during the procedure. The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the institutional review board of the National Cerebral and Cardiovascular Center (approval number: R21053-3), and all subjects gave informed consent.

PCI Procedures and RA

PCI strategies, including the decision to perform RA as well as lesion preparation, DES implantation, drug-coated balloon (DCB) use, and adjunctive therapies, were left to the discretion of the treating interventionist. RA was performed using standard techniques according to the previous consensus statements.¹⁰ RA procedure was performed using a Rotablator system (Boston Scientific, Marlborough, MA). Burr rotation speed was maintained between 140 000 and 180 000 rpm, with each ablation run limited to 20 to 30 seconds. The maximum burr size was selected by the operator based on lesion morphology and vessel size. If plaque modification with the initial burr was deemed insufficient, a larger burr was strongly recommended. Throughout the procedure, a continuous intracoronary infusion of heparin and nicorandil was administered via the burr sheath, and systemic heparin was adjusted to maintain an activated clotting time above 250 seconds. Balloon dilatation was recommended for all patients. In the RA group, balloon dilatation was performed after RA. Subsequently, PCI was completed using DES or DCB. Procedural findings included procedural time, volume of contrast agent used, achievement of final Thrombolysis in Myocardial Infarction flow grade 3, and the incidence of coronary perforation.

Acquisition of IVUS Imaging and the Definitions of Calcified Plaque Features

IVUS imaging was performed before and after PCI using either the AltaView (Terumo, Tokyo, Japan) or an OptiCross (Boston Scientific), depending on device availability and operator preference. After intracoronary administration of nitroglycerin (0.1–0.2 mg), unless contraindicated, the IVUS catheter was advanced distal to the target lesion, and imaging was obtained in a standardized fashion. All IVUS images were electronically transferred to the central imaging laboratory at the National Cerebral and Cardiovascular Center for standardized offline analysis (QIvus, Medis, Leiden, the Netherlands). CN was defined by the presence of all of the following IVUS features: (1) a convex shape of the luminal surface, (2) a convex shape of the luminal side of calcium, (3) an irregular luminal surface, and (4) an irregular leading edge of the calcification, according to previously published criteria.¹¹ The treated lesion was defined as the segment containing CN, which received DES or DCB, and both 5-mm proximal and distal segments. At baseline, IVUS measurements were obtained to assess lesion characteristics. Maximum arc of CN was defined as the greatest angular extent of calcification (in degrees) at the minimum lumen area (MLA) site within the CN segment (Figure 1). Further details regarding specific IVUS definitions and measurements are provided in Methods S2 in the Supplemental Material. For this analysis, 2 additional morphological features were assessed. First, adjacent circumferential calcification was defined as calcification with a 360° arc within this segment (Figure 1). Second, contralateral calcification was defined using a quadrant-based analysis at the MLA cross-section. The lumen cross-section was divided into 4 quadrants, with the quadrant containing the CN designated as the index quadrant. The presence of any calcification in the directly opposite quadrant was defined as contralateral calcification (Figure 1). IVUS image interpretation was independently conducted by 2 experienced cardiologists (N.Y. and M.F.) who were blinded to all patient characteristics and clinical outcomes. Interobserver and intraobserver agreement showed excellent reproducibility for identification of both CN (κ=0.83 and κ=0.84, respectively) and contralateral calcification (κ=0.85 and κ=0.90, respectively).

Figure 1.

Definitions of key intravascular ultrasound (IVUS)–derived morphological features. This figure illustrates the measurement methods and definitions for the key IVUS-based morphological features used in this study. A, Measurement of the maximum arc of the calcified nodule (CN; indicated by the asterisk), defined as the angular extent of calcification (indicated by the area between the dotted lines) at the site of the minimum lumen area. B, A longitudinal IVUS image of a lesion containing CN, indicating the specific locations of the cross-sectional images shown in C and D with yellow dots. C, Representative cross-sectional images at the CN site defining the presence or absence of contralateral calcification. Contralateral calcification was defined as the presence of calcification in the quadrant opposite to the CN. C′, A case with contralateral calcification, whereas C″ shows a case without it. D, Representative images in the adjacent lesion segment of CN defining the presence or absence of circumferential calcification. Adjacent circumferential calcification was defined as a calcification arc of 360°. D′, A case with circumferential calcification, whereas D″ shows a case without it.

Quantitative Coronary Angiography

Quantitative and qualitative coronary angiographic analyses were independently performed by 2 cardiologists (N.Y. and M.F.) who were blinded to both clinical data and IVUS findings. Details regarding the analyses are provided in Methods S3 in the Supplemental Material.

Outcomes and Clinical Follow-up

The primary outcome of this study was the occurrence of TLR during the follow-up period. TLR was defined as any repeat PCI or coronary artery bypass grafting performed due to restenosis at the previously defined segment.¹² Secondary outcomes were assessed, including all-cause death, cardiac death, target-vessel myocardial infarction, target-vessel revascularization, and target-vessel failure. Target-vessel failure was defined as a composite of cardiac death, target-vessel myocardial infarction, or target-vessel revascularization. All end points were defined according to the Academic Research Consortium-2 consensus document.¹² Patients were monitored through regular follow-up visits at outpatient clinics of each participating institution, by general practitioners, or both, every 1 to 3 months, and those who were lost to follow-up were censored at the date of their last known contact. Details regarding noninvasive testing and indications for revascularization are provided in Methods S4 in the Supplemental Material.

Statistical Analysis

Continuous variables were expressed as the mean±SD and compared using the t test if normally distributed. Non-normally distributed variables were presented as medians with interquartile ranges. Categorical variables were compared using either the Fisher exact test or χ² test, as appropriate. Patients were divided into 2 groups based on RA use, and baseline clinical characteristics, procedural details, and IVUS findings were compared between the 2 groups. The cumulative incidence of TLR was estimated using the Kaplan-Meier method, and differences between the groups were assessed with the log-rank test. In this analysis, the proportional hazards assumption was tested using the Schoenfeld residuals test. To assess the association between RA and TLR, we conducted Cox proportional hazard analysis and presented the results as hazard ratios (HRs) and 95% CIs. To separately evaluate clinical and IVUS-derived morphological predictors while mitigating the risk of overfitting, we constructed multiple multivariable models. Model 1 focused on clinical characteristics, adjusting for age, sex, diabetes,¹³ dyslipidemia,¹⁴ and hemodialysis,¹⁵ which were selected based on established or potential association with TLR from previous studies and clinical significance. Model 2 assessed the association between RA and TLR, adjusting for the impact of IVUS-derived morphological features. For the final integrated model (model 3), least absolute shrinkage and selection operator–based Cox penalized regression was conducted as a variable selection step before Cox proportional hazards regression. The selected covariates and female sex, included due to clinical importance, were used as adjustment covariates. This model construction, incorporating RA use, adheres to established statistical principles regarding the rule of thumb.¹⁶ In addition, prespecified subgroup analyses were conducted to evaluate whether the association between RA use and TLR differed across clinical and IVUS-derived characteristics. For IVUS-derived features, specifically the maximum arc of CN and CN length and lumen area stenosis, patients were divided into 2 groups based on the median value. The interaction between RA use and each subgroup variable was assessed using Cox regression models with interaction terms. Forest plots were generated to visually present HRs and CIs across subgroups. To further examine the robustness of the association regarding covariate balance, we performed a 1:1 propensity score–matched analysis (Methods S5 in the Supplemental Material). To account for potential clustering of patients within the participating centers, the study site was included as a fixed effect in both the multivariable regression models and the analysis of the propensity score–matched cohort. All statistical tests were 2-sided, and a P<0.05 was considered statistically significant. Statistical analyses were performed using SPSS (IBM Corp, Armonk, NY) and R version 4.4.1.

Results

Study Population and Baseline Clinical Characteristics

From April 2008 to December 2024, a total of 348 patients with coronary artery disease underwent IVUS-guided PCI for culprit lesions with CN (Figure S1). Among them, we excluded patients with in-stent restenosis (n=60), treated using alternative debulking devices—orbital atherectomy (n=25), excimer laser coronary atherectomy (n=5), or directional coronary atherectomy (n=2), lesions in which no device could be advanced without RA (n=23), and those with poor IVUS image quality (n=24). Thus, 209 patients with de novo CNs were included in this analysis. Among them, 79 patients (37.8%) underwent PCI with RA (RA group), whereas 130 (62.2%) were treated without RA (non-RA group). Baseline characteristics are shown in Table 1. There were no significant differences in baseline clinical characteristics and coronary risk factors between the 2 groups. Patients receiving RA were less likely to present with acute coronary syndrome. With regard to guideline-recommended medical therapies, a lower frequency of statin use was observed in patients receiving RA (Table 1).

Table 1.

Baseline Clinical Demographics

Angiographic and Procedural Findings

Angiographic characteristics and procedural details are shown in Table 2. The right coronary artery was the most common location of CNs (57.4%), followed by the left anterior descending artery (16.7%), left circumflex artery (13.4%), and left main trunk (12.0%). There were no significant differences in CN distribution between the RA and non-RA groups. Severe angiographic calcification was more frequently observed in the RA group compared with the non-RA group (67.1% versus 35.4%; P<0.001).

Table 2.

Angiographic Characteristics and Procedural Details.

In the RA group, 28 patients (35.5%) were treated with a burr ≤1.5 mm, 22 patients (27.8%) with a 1.75-mm burr, and 29 patients (36.7%) with a burr ≥2.0 mm. After the RA procedure, all of the patients in the RA groups received balloon angioplasty to further dilate CN. There were no differences in preparation balloon type or size between the RA and non-RA groups. Intravascular lithotripsy (IVL) was used in 7 cases, all of which were in the non-RA group. After the completion of these procedures, DES and DCB were used in 83.3% (P=0.29) and 16.7% of the entire subjects, respectively. The stent diameter was 3.5 (3.0–3.5) mm in the RA group and 3.25 (3.0–3.5) mm in the non-RA group (P=0.09). There were trends toward larger sizes of DES and DCB in the RA group, but these comparisons did not meet statistical significance (DES: P=0.09, DCB: P=0.07; Table 2). A summary of procedural findings is available in Table S1. Procedural time was longer in the RA group than in the non-RA group, while contrast volume was comparable between the 2 groups. The final Thrombolysis in Myocardial Infarction 3 flow grade was similar between the groups (97.5% versus 99.2%; P=0.30). Coronary perforation occurred in only 1 case in the RA group, with no statistically significant difference between the groups.

IVUS Findings and Calcified Plaque Features

Preprocedural IVUS analysis revealed that lesions in the RA group had more severe calcific morphology (Table 3), including a larger maximum calcium arc at the entire lesion and greater lumen area stenosis (both P<0.01), as well as a larger maximum arc of CN (P=0.011). Adjacent circumferential calcification and contralateral calcification were significantly more prevalent in the RA group (31.6% versus 19.2%, P=0.041, 44.3% versus 23.8%, P=0.002). There were no significant differences in calcium or CN length, or in the preprocedural MLA at both sites.

Table 3.

Characteristics of IVUS Results

Postprocedural IVUS showed that the RA group achieved a significantly greater final MLA and acute lumen gain at both the entire lesion and the CN site (all P<0.01). Despite the greater acute gain, final stent expansion and eccentricity were comparable between the groups.

Clinical Outcomes and Predictors of TLR

During the follow-up period (median, 2.1 years, [interquartile range, 0.4–4.9]), TLR was required in 20 of 79 patients with RA use (25.3%) and 41 of 130 patients without RA use (31.5%), respectively. The cumulative incidence of TLR was significantly lower in the RA group compared with the non-RA group during the follow-up period (log-rank P=0.040; Figure S2). The details of the secondary outcomes are shown in Table 4. After adjustment for baseline clinical characteristics (model 1), RA use remained independently associated with reduced TLR (adjusted HR, 0.48 [95% CI, 0.28–0.84]; P=0.010; Table 5). In a separate model incorporating IVUS-derived morphological features (model 2), the protective association of RA was even more pronounced (adjusted HR, 0.33 [95% CI, 0.18–0.60]; P<0.001). In addition, greater lumen area stenosis, longer CN length, adjacent circumferential calcification, and a smaller final MLA were identified as independent predictors. Finally, in the integrated model (model 3), RA use remained independently associated with a lower risk of TLR (HR, 0.34 [95% CI, 0.19–0.62]; P<0.001). A propensity score–matched analysis yielded 64 matched pairs with well-balanced baseline characteristics (Table S2; Figure S3). In this cohort, although postprocedural IVUS measurements were comparable (Table S3), RA use remained significantly associated with a lower risk of TLR (HR, 0.45 [95% CI, 0.24–0.83]; P=0.011).

Table 4.

Clinical Outcomes During Follow-Up

Table 5.

Univariate and Multivariable Cox Regression Analyses of Predictors for TLR

Subgroup Analysis of IVUS-Derived Calcification Features Favorable for RA

In the subgroup analysis, RA use was associated with favorable TLR rates across most subgroups (Figure 2). In subgroups characterized by more severe calcification—such as those with hemodialysis, a large maximum arc of CN, long CN length, and a high percent of lumen area stenosis—RA was consistently associated with lower HRs. Notably, the benefit of RA in reducing TLR was more pronounced in patients with contralateral calcification (8.6% versus 51.6%, P<0.001), whereas no significant difference was observed in those without it (38.6% versus 25.3%, P=0.11). The interaction between RA effect and contralateral calcification was statistically significant (P<0.001; Figure 2).

Figure 2.

Forest plots of subgroup analyses for target lesion revascularization (TLR). The forest plot illustrates the hazard ratios (HRs) for TLR, comparing patients treated with rotational atherectomy (RA) vs those without RA across various prespecified subgroups. Each square denotes the HR, with the corresponding 95% CI shown as a horizontal line. P values for interaction (P_interaction) were calculated to determine if the effect of RA on TLR risk differed significantly across subgroups. The strongest interaction was found with the presence of contralateral calcification to calcified nodule (CN; P_interaction <0.001). No other significant interactions were detected across the remaining subgroups. ACS indicates acute coronary syndrome; DCB, drug-coated balloon; DES, drug-eluting stent; and RCA, right coronary artery.

Discussion

This study provides 3 key insights into the outcomes and therapeutic effect of RA for CNs. First, RA use was independently associated with a significantly lower risk of TLR over the long-term follow-up. Second, independent of RA use, specific morphological features—greater lumen area stenosis, longer CN length, smaller final MLA, and the presence of adjacent circumferential calcification—were identified as powerful predictors of increased TLR risk. Third, the therapeutic benefit of RA was not uniform; a significant treatment interaction was observed, with the effect being most pronounced in patients with contralateral calcification. These findings underscore the importance of a morphology-guided treatment strategy for CNs, helping to identify both lesions at high risk of TLR and those most likely to benefit from RA.

Efficacy of RA for CNs

In this study, RA was independently associated with a significant reduction in long-term TLR for CNs. Our positive findings contrast with a previous single-center, propensity score–matched study that reported no significant clinical benefit of RA for CNs.⁸ There may be 2 key differences to explain this discrepancy. First, our study provides a more focused assessment of RA’s therapeutic effect by excluding cases where RA was indispensable for device passage, a potential confounder not addressed previously. Second, our study used larger burr sizes (≥1.75 mm in 65.5% of cases), which likely enabled more effective plaque modification compared with the predominantly small burrs (≤1.5 mm in 85% of cases) used in the prior study. A ROTA.shock trial (Comparison of Coronary Lithoplasty and Rotablation for the Interventional Treatment of Severely Calcified Coronary Stenoses) substudy suggested that RA did not significantly reduce the plaque volume of CNs, with lumen gain achieved primarily by plaque displacement during stenting.¹⁷ This implies that the primary role of RA is plaque modification to facilitate expansion, rather than debulking itself, and that an adequate ablation strategy is, therefore, essential to alter outcomes. This interpretation is strongly supported by our IVUS findings: while the preprocedural MLA, measured for both the entire lesion and the CN site, was comparable between the groups, the RA group achieved a significantly larger final MLA at both sites. Interestingly, the benefit of RA on long-term TLR was independent of the final MLA, suggesting mechanisms beyond acute luminal gain. Severe calcification is known to predict adverse outcomes, including TLR after DES or DCB.^18,19 Although this is often attributed to stent underexpansion, preclinical evidence suggests other potential mechanisms: one ex vivo study showed that severe calcification impedes antiproliferative drug delivery, and that atherectomy can modify the barrier to enhance drug absorption and penetration.²⁰ Furthermore, Torii et al²¹ reported that stenting severely calcified lesions is associated with significantly deeper vessel injury (medial tears) than in nonseverely calcified lesions, a factor linked to adverse vascular responses. Plaque modification with atherectomy before stent expansion may mitigate this injury. Therefore, RA might improve drug delivery and vascular healing by reducing vessel injury, contributing to its association with lower long-term TLR, independent of acute lumen gain. This overall finding of RA’s benefit is consistent with another retrospective analysis, which also concluded that not using RA was an independent predictor of restenosis.⁷

Predictors of Long-Term TLR

Our findings suggest that overall calcification severity—reflecting an advanced lesion phenotype—is a major determinant of TLR risk. Specifically, greater lumen area stenosis, longer CN length, and adjacent circumferential calcification were all strong, independent predictors of TLR. These results underscore the importance of calcific burden not only within the CN itself but also in the surrounding plaque. This aligns with our previous work, which demonstrated that adjacent calcification is a key driver of TLR.⁹ In univariate analysis, a large maximum CN arc is also associated with increased TLR risk, consistent with recent optical coherence tomography findings.²² Importantly, our analysis also highlighted the prognostic value of the postprocedural result. A larger final MLA remained an independent protective factor, highlighting the importance of achieving sufficient final lumen area to maintain long-term patency. The fact that the RA group achieved a significantly larger final MLA supports the role of effective lesion preparation in a favorable clinical outcome. Specifically, our procedural strategy—with an optimal burr-to-artery ratio (0.56) and high rate of burr upsizing (36.7%)—likely achieved more effective plaque modification compared with previous conservative approaches, which may explain our favorable outcomes.

Contralateral Calcification as a Key Modifier of RA Efficacy

The most novel finding of this study is the strong treatment interaction observed between RA use and the contralateral calcification. This indicates that the benefit of RA is not uniform across all lesions but is highly influenced by specific lesion morphology. Our results suggest that contralateral calcification functions as an effect modifier rather than a direct risk factor for TLR. The presence of contralateral calcification may lead to suboptimal luminal expansion if lesion preparation is inadequate. However, this calcification can act as rigid structural support, serving as a backstop during RA. This improves guidewire bias and provides stability that allows the burr to ablate the CN more effectively. Without this support, the vessel may shift or deflect during atherectomy, resulting in inefficient plaque modification. Indeed, RA achieved significantly more effective plaque modification, measured by acute luminal gain from the procedure itself, specifically in patients with contralateral calcification compared with those without after adjusting for confounders (Table S4). This finding objectively supports this backstop hypothesis. In our registry, operators were more likely to perform RA in lesions with contralateral calcification, likely recognizing its technical advantage. Furthermore, the presence of contralateral calcification appears to characterize a particularly challenging lesion subset for conventional therapy. The prohibitively high TLR rate (51.6%) observed in patients with this feature who were treated without RA confirms the high-risk nature when inadequately modified. Therefore, contralateral calcification may serve as a practical biomarker for identifying patients most likely to benefit from RA, enabling a more targeted and effective treatment strategy (Figure S4).

Clinical Implications and Future Directions

Our findings do not support the routine use of RA for all CNs. Rather, treatment strategies should be tailored to the individual patient’s clinical condition and lesion morphology, with consideration of alternative devices when appropriate. Nevertheless, our findings offer a practical framework for the application of RA in the management of CNs, shifting the paradigm from a general debulking strategy to a more selective, morphology-guided approach. The significant interaction observed with contralateral calcification suggests that RA should be strongly considered for patients with this specific anatomic feature, as they are most likely to derive a substantial long-term benefit. Furthermore, our findings also indicate that a high lumen area stenosis, long CN length, and adjacent circumferential calcification are strong independent predictors of TLR. These features should prompt caution, as they identify a subset of patients in whom RA alone may be insufficient. In such cases, more advanced and aggressive lesion preparation may be warranted. Moreover, because absolute TLR rates remain substantial even in favorable subgroups, future studies are needed to evaluate whether intensified plaque modification, such as a combined shave and shock approach using RA and IVL,^17,23 can improve long-term outcomes in this challenging population. Future observational or randomized studies could therefore compare patients with high-risk CN features treated with an RA-only strategy versus a combined strategy (eg, the Dual-Prep registry [Atherectomy Devices and Intravasular Lithotripsy for the Preparation of Heavy Calcified Coronary Lesions Registry] [URL: https://jrct.mhlw.go.jp/; Unique identifier: jRCT1032230384] or the NODULE-SHOCK trial (Intravascular Lithotripsy With or Without Rotational Atherectomy for Coronary Calcified Nodule Treatment) [URL: https://www.clinicaltrials.gov; Unique identifier: NCT07000045]).

Limitations

This study has several limitations. First, this was a retrospective analysis from a multicenter registry and is subject to potential selection bias, including confounding by indication. The decision to use RA was at the operator’s discretion. Although we attempted to mitigate this treatment selection bias using a propensity score analysis, specific procedural techniques—including the choice of RA burr size and rotational speed, as well as the type of balloon used for subsequent lesion preparation—also varied by operator and were not adjusted for, which may have introduced bias despite statistical adjustments. Second, our diagnosis of CNs was based on IVUS, not optical coherence tomography. Consequently, we could not differentiate between eruptive and noneruptive CNs, a distinction with known prognostic importance. While optical coherence tomography provides superior resolution for such characterization, its use was limited in our cohort due to anatomic and clinical constraints. CNs were frequently located in ostial lesions (33.5%), and a high proportion of patients had chronic kidney disease (70.8%). In these contexts, IVUS is often preferred to ensure adequate imaging of ostial segments and to avoid the risks associated with contrast administration. Third, the number of cases treated with IVL was small. Specifically, IVL cases (n=7, 3.3%) were included in the non-RA group. This was partly due to the fact that during the study period, the combined use of RA and IVL was not reimbursed by the national health insurance system in Japan, limiting its application as a bailout or combination strategy. Although the low number of IVL cases likely had a minimal effect on the overall comparison, the inclusion of this potent device only in the non-RA group remains a specific limitation. This constrains the generalizability of our findings regarding modern multimodal lesion preparation using IVL. Fourth, the sample size, although substantial for a study on CNs, may have limited the statistical power for subgroup analyses. Finally, our study cohort consisted entirely of Japanese patients treated at high-volume centers. Therefore, the generalizability of our findings to other ethnic populations or to different practice settings—where RA techniques and burr sizing strategies may vary—requires further investigation.

Conclusions

In this multicenter registry of patients with CNs, RA was independently associated with a significant long-term reduction in TLR. This therapeutic benefit was most pronounced in CNs with contralateral calcification. In contrast, markers of lesion severity, such as longer CN length, greater lumen area stenosis, and adjacent circumferential calcification, remained strong independent predictors of TLR. These findings support a morphology-guided approach to CN treatment, allowing interventionists to better stratify lesion-specific risk and identify patients most likely to benefit from atherectomy. Prospective, randomized trials are warranted to confirm these observational findings.

ARTICLE INFORMATION

Acknowledgments

The authors are deeply grateful to the attending physicians and co-medical staff members across all participating U-SCAN registry (Coronary Intravascular Ultrasound for Calcified Nodule) sites for their invaluable collaboration and support during the percutaneous coronary intervention procedures. We offer particular gratitude to the clinical engineers: Sayaka Watanabe, Ryoma Saga, Shintaro Kobayashi, and Tomoyu Kondo. We also extend our sincere appreciation to all investigators, clinical research coordinators—especially Yuko Yoshioka—data managers, and the patients who participated in the U-SCAN registry.

Disclosures

Dr Yabumoto has received honoraria from Abbott Medical and Amgen. Dr Murai has received honoraria from Abbott Medical, Terumo, Amgen, Astellas, Zeon Medical, and Boehringer, as well as support for attending meetings from OrbusNeich. Dr Tsujita has received research support from Abbott Medical, Alexion Pharma, Bayer Yakuhin, Biotronik, Boston Scientific, Daiichi Sankyo, EA Pharma, Fides-one, Fukuda Denshi, GE HealthCare, GM Medical, ICON Clinical Research, ITI, Japan Lifeline, Kaneka Medix, Medtronic, Mochida Pharmaceutical, Nipro, Novo Nordisk Pharma, Otsuka Medical Devices, Philips Japan, PPD-SNBL, Roche Diagnostics, and Terumo; and has received honoraria from Abbott Medical, Amgen, Bayer Yakuhin, Daiichi Sankyo, Kowa Pharmaceutical, Mochida Pharmaceutical, MSD, Novartis Pharma, Otsuka Pharmaceutical, Pfizer, Takeda Pharmaceutical, and Novo Nordisk Pharma. Dr Kataoka has received research support from Kowa, Nipro, and Abbott and honoraria from Nipro, Abbott Medical, Kowa, Amgen, Sanofi, Astellas, Takeda, and Daiichi Sankyo. The other authors report no conflicts.

Supplemental Material

Supplemental Methods

Tables S1–S4

Figures S1–S4

References 24–26

Funding Statement

Dr Yabumoto received research grants from the Cardiovascular Research Fund, Japan Arteriosclerosis Prevention Fund, Konica Minolta Science and Technology Foundation, and Terumo Life Science Foundation.