Comparison of Angiographic Result and Long-Term Outcome in Patients With In-Stent Restenosis Treated With Cutting Balloon or With Scoring Balloon Angioplasty

© 2024 HMP Global. All Rights Reserved.
Any views and opinions expressed are those of the author(s) and/or participants and do not necessarily reflect the views, policy, or position of the Journal of Invasive Cardiology or HMP Global, their employees, and affiliates. 

Background. Lesion preparation with a cutting balloon (CB) or scoring balloon (SB) is often used in patients with in-stent restenosis (ISR). However, there are no comparative studies.

Methods. We analyzed 81 patients (CB group: n = 38; SB group: n = 43) who had a calcified ISR from November 2019 to September 2021. The primary endpoint was strategy success (< 20% residual stenosis); the secondary endpoints were major adverse cardiovascular events during the 1-year follow-up. Quantitative coronary angiography was performed to evaluate the strategy success.

Results. The patients in the CB group were more likely to have a severely calcified ISR (P = .001) and multiple stent layers (P = .001). A total of 64 patients (79.0%) reached the primary endpoint. Residual stenosis greater than 20% was more common in the CB group (39.5% vs 4.7%; P = .001). In the multivariate analysis, an effect of the intervention group on the achievement of the primary endpoint could be excluded (estimate 1.06; standard error 1.07; P = .322). The time interval of stent implantation prior to CB/SB (P = .007) and severe calcified ISR (P = .009) had a negative impact on reaching the primary endpoint. During the follow-up, there were no differences in rates of cardiac death (CB 2.5% vs SB 1.2%; P = .598), acute myocardial infarction (CB 0% vs SB 4.9%; P = .119), and target lesion failure (CB 3.7% vs SB 12.3%; P = .074).

Conclusions. In our cohort, multivariate analysis showed that lesion preparation with CB or SB must be considered equivalent in terms of angiographic results. Factors like severe calcified ISR and the time interval of prior stent implantation negatively influenced the angiographic outcome.

In-stent restenosis (ISR) is characterized by a significant reduction of the lumen diameter within a stent. ISR occurs in ­­5% to 10% of cases and is the most common cause of stent failure.^1,2 The mechanism of ISR depends on various factors, including factors related to the initial implantation, such as under expansion, under sizing, fracture/gap, and stent type.² In addition, intra-stent factors such as neointimal hyperplasia, neo atherosclerosis, and calcification may lead to ISR. Furthermore, multiple stent layers, vessel calcification or calcified nodules, the vessel size, and residual plaque burden increase are attributed to the development of an ISR.² In general, there is no significant difference in lesion preparation of a de novo stenosis compared to an ISR. However, this topic has received limited study.

Cutting balloons (CB) and scoring balloons (SB) are commonly used for the preparation of a calcified lesion.³ In most studies, these 2 balloon types are summarized under the term modified balloons (MB) and the guidelines make no clear recommendation as to when they should be used.^4,5 Furthermore, the 2 balloon types differ in their design. The principle of these devices is the focused transmission of force from the balloon to the vessel wall. This is done by microsurgical blades (CB), or plastic threads integrated in the balloon or scoring elements (SB). These blades or scoring elements are pressed into the vessel wall during balloon inflation but are folded in when deflated. The blades (CB) are placed longitudinally on the surface of the balloon, resulting in controlled longitudinal incisions in the dilated target segment. In the SB, the scoring elements are embedded in the balloon folds and score the plaque circumferentially when the balloon is inflated.

There is a lack of comparative randomized studies investigating the effectiveness and safety of the 2 balloon types. MB balloons help prevent balloon slippage, but their higher profile may limit deliverability and the efficacy in calcified lesions is controversial.⁴ A recently published meta-analysis showed that in patients treated with an MB prior to stent implantation for severe coronary artery disease, no improvement in clinical or imaging outcomes compared to control therapy.⁴ However, these studies only refer to de novo stenoses.^6-11 One trial demonstrated lower rates of restenosis with the use of an SB compared with a conventional balloon angioplasty prior to a drug-coated balloon (DCB) in the treatment of ISR.¹² However, there are no comparative data in the lesion preparation of an ISR with either a CB or SB. Therefore, this study is the first to investigate the efficacy and safety of CB vs SB in calcified ISR lesion preparation.

Patient selection and study design. The present study is a non-randomized prospective registry. Patient recruitment took place in the Heart Centre Trier and the German Heart Centre, Technical University, Munich between November 2019 and September 2021. During this period, 3244 patients underwent percutaneous coronary intervention (PCI) in both centers. A total of 561 lesions (17.3%) were ISR. In 117 patients (3.6%), the ISR was graded as calcified. In 81 patients (2.5%), angioplasty was performed using CB or SB angioplasty.  All these patients had angiographically calcified ISR (Figure 1) and the calcification was graded into moderate or severe, as previously described.^13-15 It was recommended to perform intravascular imaging. The decision whether and which intravascular imaging method was used was the responsibility of the operator.

The patients were divided into the following groups depending on the use of the balloon-type: CB (n = 38; WOLVERINE Cutting Balloon [Boston Scientific]) vs SB (n = 43; NSE Alpha [B. Braun]). The selection of the CB or SB diameter was based on the size of the previously implanted stent (1:1 ratio). The operators were allowed to decide for themselves whether to use a CB or SB. All other decisions, such as pre-dilatation, post-dilatation, and use of a DES or DCB, were also the responsibility of the interventional cardiologist. Informed consent was obtained from each patient, and the study was approved by ethics committees of the participating centers (No. 2019-14544).  

Quantitative coronary angiography. We performed offline quantitative coronary angiography (QCA) in all patients. The QCA was evaluated as previously described.^13,16 In brief, the analysis was performed before CB or SB dilatation and at the end of the procedure (Figure 1A-C). The stenoses were categorized as eccentric or concentric. The severity of calcification was assessed using angiographic criteria as previously described.^13,16 In brief, a moderate calcification was assumed if the calcification was only seen during the cardiac cycle.  If the calcification was visible even without cardiac motion, it was classified as severe. Treatment success (primary endpoint) was achieved if a residual stenosis of less than 20% was present in the QCA. Strategy failure was present if the residual stenosis in the QCA was greater than 20% (Figure 1D).

Follow-up. The following periprocedural events were recorded: periinterventional cardiopulmonary resuscitation and acute myocardial infarction (AMI). We also analyzed the following potential device-related events: coronary artery dissection during the index procedure, coronary artery perforation after CB or SB, thrombus formation during PCI, and acute stent-thrombosis. The diagnosis or exclusion of a peri-interventional AMI was based on the fourth universal definition of MI.¹⁷ All patients underwent a telephone interview after 365 days.
 

Endpoints. The primary endpoint was strategy success, which was defined as successful DES/DCB delivery and expansion with less than 20% residual stenosis of the treated ISR. Secondary endpoints were major adverse cardiovascular events (MACE), which were defined as death, AMI, and target vessel revascularization (TVR), which included target vessel failure. 

Statistical analysis. Statistical analysis was performed as described previously.¹⁶ Categorical variables are presented as n (%). Continuous variables are summarized as mean ± SD or as the median with interquartile range (IQR). 

Prior to the analysis, the data was checked for a possible bias and the need to carry out propensity score matching. The analysis carried out and validated with bootstrapping did not reveal any relevant bias. A separate multivariate regression analysis was developed to further verify the parameters that influence the primary endpoint. In addition to the p-values, an estimate value was obtained for each predictor, interpretable as a log odds ratio (OR = estimate value e^x). The following variables were tested: intervention group, severe calcification, multiple stent layers, lesion length, time interval of stent implantation prior to CB or SB, pre-dilatation balloon type, post-dilatation balloon type, post-dilatation with a super high-pressure balloon, post-dilatation pressure, diameter of DES/DCB, length of CB/SB, fluoroscopy time, DES implantation, and DCB usage.

The variables of the secondary endpoint were analyzed using univariate Cox proportional hazard ratio (HR) and Kaplan-Meier analysis. R 4.2.2 software (R Foundation for Statistical Computing) was used for the statistical analysis.

Baseline and angiographic characteristics and procedural data. A total of 81 patients were included in the final analysis. Of these, 38 patients were in the CB group and 43 patients were in the SB group. At the time of admission, no significant differences in baseline characteristics were observed between the groups (Table 1). Further, there were no differences in the clinical presentation between the groups.

The different lesion characteristics are shown in Table 2. A culprit lesion of the left circumflex artery was less common in the CB group (P = .005). Patients in the CB group were more likely to have a severely calcified ISR (P = .001) and multiple stent layers (P = .001). In the CB group, the interval between first stent implantation and ISR was significantly longer (P = .001). No further significant differences in lesion characteristics were detected (Table 2).

Procedural data are shown in Table 3. Pre-dilatation with a semi-compliant balloon was less frequent (P = .015) and pre-dilatation with a non-compliant balloon was more frequent (P = .015) in the CB group. Post-dilatation with a semi-compliant balloon was also less frequent in the CB group (P = .028). Post-dilatation pressures were higher in the CB group (P = .004). Due to the different balloon lengths available, the balloons in the CB group were longer than those in the SB group (P = .023) and larger balloon diameters were used in the CB group (P = .003). DES implantation was more frequent in the CB group (P < .001) and a lower rate of DCBs was observed in the CB group (P < .001). The diameter of DES/DCB was smaller in the CB group (P = .007). However, the length of the DES/DCB was longer in the CB group (P = .008).

Procedural and in-hospital events. No MACE events occurred during the hospital stay. There were also no periprocedural complications.

Primary efficacy endpoint. A total of 64 patients (79.0%) reached the primary endpoint. Overall, 23 patients (28.4%) in the CB group and 41 patients (50.6%) in the SB group met the primary endpoint (P < .001; Table 3). Due to the differences in lesion characteristics and lesion preparation as described above, a multivariate model was developed. Multivariate regression analysis was used to further verify the parameters that influence the primary endpoint and to reduce the selection bias. The data of the multivariate regression analysis are shown in Table 4. Even though there was a significant difference in the descriptive statistics between the groups regarding the achievement of the primary endpoint, an effect of the intervention group on the primary endpoint could be excluded (estimate 1.06; standard error [SE] 1.07; P = .322) in the multivariate analysis. The presence of a severely calcified ISR (P = .009), a lesion length between 10 mm and 20 mm (P = .027), and a longer time interval of stent implantation prior to CB/SB (P = .007) had a negative impact on the primary endpoint. Factors like higher post-dilatation pressure (P = .046) and longer procedure time as well as fluoroscopy time (P = .003, respectively P = .009) also had a negative impact on the primary endpoint. The only factor that had a positive impact on the primary endpoint was the DES/DCB diameter in our cohort (P = .016). The use of a DCB instead of a DES was considered to have equal probability of reaching the primary endpoint (P = .237).

Secondary endpoints and long-term follow-up. Long-term follow-up was available for all 81 patients (Table 5). There were no significant differences in cardiovascular mortality rate (CB n = 2 [2.5%] vs SB n = 1 [1.2%]; P = .598) and in rate of AMI (CB n = 0 [0%] vs SB n = 4 [4.9%]; P = .119). No significant difference was observed in the rate of TVR between the groups (n = 3 [3.7%] vs n = 10 [12.3%]; P = .074). Univariate regression analysis could also exclude a significant effect of the intervention group on the TVR rate (HR 3.19; 95% CI, 0.89-11.58; P = .078).  Multiple stent layers (HR 1.08; 95% CI 0.33-3.5; P = .901), severe calcification (HR 0.804; 95% CI, 0.221-2.92; P = .741), and the time interval between prior stent implantation and the index procedure (HR 1; 95% CI, 0.99-1.0; P = .192) also had no effect on the TVR rate. In addition, the use of a DCB or DES had no effect on the TVR rate in our cohort (DCB: P = .952; DES: P = .875). The Kaplan-Meier analysis of freedom from TVR is shown in Figure 2.

Interventional treatment of ISR is still a challenge and there is a lack of recommendation in the guidelines on how to treat it.⁵ MB is often used in lesion preparation prior to DESs or DCBs in the treatment of ISR even though the evidence for its use is still limited.^{2,5,10,13,18,19} In particular, most studies do not differentiate between CB and SB, even though they are designed differently. This registry is the first to investigate the effect of CB or SB dilation in the treatment of ISR due to in-stent calcification.  

Our results can be summarized as follows: (1) the multivariate regression analysis could exclude an effect of the intervention group on the primary endpoint, although in the descriptive statistical analysis, the number of patients who met the primary endpoint was significantly less in the CB group; (2) no device-related procedural complications or in-hospital adverse events occurred in either of the groups; and (3) there were low rates of MACE in both groups without significant difference during the follow-up.

Procedural success and mechanism of action. The overall success rate was 79%. Although lesion preparation with MB is commonly used for calcified stenoses, there is limited data investigating its use in ISR and, to our knowledge, there have been no previous studies that included cardiovascular endpoints. Our overall success rate did not differ from that of other studies where rates are reported as between 81.0% to 81.5%.^10-12At the same time, lower rates have been reported in the RESCUT trial. Here, 428 patients with ISR were randomized to treatment with either CB or conventional balloon angioplasty. The authors report a binary restenosis rate of 30% in patients treated with CB and 31% in patients with conventional PCI.¹⁸ However, this study is almost 20 years old, and the lesion preparation devices have improved since then.

In our multivariate analysis, the influence of CB or SB on the primary endpoint could be excluded. The failure of strategy depended more on the severity of intraluminal ISR calcification, the length of the lesion, and the time interval between the original stent implantation and the subsequent intervention performed. This aligns with recent results where lesion length and calcification were predictors of angiographic ISR over the time.^20,21 In particular, a recently published study showed that a mechanical cause (such as severe calcification) leads to a higher ISR rate.²² This is also confirmed by a study investigating the associated factors of ISR with optical coherence tomography.¹⁹ Here, old stent under expansion, multiple layers of old stent, maximum calcium angle greater than 180°, and maximum calcium thickness greater than 0.5 mm were independently associated with new stent under expansion.¹⁹

A meta-analysis could demonstrate that lesion preparation with an MB before stenting in patients with severe calcified coronary arteries did not improve the clinical or imaging outcomes compared with control therapy.⁴ However, this study did not investigate patients with ISR.

An advantage of using a CB or SB over conventional balloon angioplasty is the prevention of balloon slippage (“watermelon seeding” [WMS]) during treatment of ISR. In a study by Alfonso et al, the authors report that patients with WMS had more severe and diffuse ISR lesions.^2,23 They observed that WMS led to a higher frequency of balloon inflations and longer total inflation times. Furthermore, these patients had more residual dissections and a higher rate of crossover to stenting, as well as poorer acute results as demonstrated by lower minimal lumen diameters after PCI compared to ISR patients without WMS.^2,23 This phenomenon potentially complicates treatment of ISR. Factors associated with this technical problem are long and severe lesions. The occurrence of the WMS effect during stent deployment has not been reported so far.^2,23 However, because we used either a CB or SB, the WMS phenomenon was not present in our cohort.

A potential advantage of the SB over the CB could be better crossability. In our cohort, patients in the CB group had a significantly longer procedure time and a higher rate of pre-dilatation with a non-compliant balloon, which could support this assumption. However, there is only 1 study in the literature that deals with crossability. Contrary to our expectations, this study reports an improved crossability with novel CB vs SB in the treatment of calcified stenoses.³ However, this was not investigated in ISR and thus cannot be accurately compared to our data.

Multivariate regression analysis revealed that a larger DCB/DES diameter led to a higher probability of a residual stenosis less than 20%. This can be explained by the lesion preparation: successful plaque modification allows larger DCB/DES diameters to be used than if there is still a stenosis after lesion preparation. Furthermore, higher post-dilatation pressure, long procedure time, and long fluoroscopy time negatively influenced the probability of reaching the primary endpoint. These factors should be considered as interventional cardiologists work to improve the primary interventional outcome.

Procedural safety and follow-up analyses. Low rates of periprocedural complications as well as no differences in the secondary endpoints during the follow-up were observed in both groups. The described rates of cardiac death during follow-up range from 1.0% to 2.2%,^19,24 which is comparable to our cohort (CB 2.5% vs SB 1.2%). The reported rate of AMI during the follow-up is also comparable to our cohort. This varies from 1.9% to 4.4%,^12,19,24 compared to our rate of 0% in the CB and 4.9% in the SB group. The TVR rate in existing literature also did not differ from ours, ranging from 13.3% to 16.2%; in our cohort, CB was 3.7% and SB was 12.3%, with an overall rate of 16%.

A potential advantage of lesion preparation with MB could be an improvement in the anti-restenotic efficiency of the DCB.² This was demonstrated in the ISAR-4 study where the neointimal modification with SB improves the anti-restenotic efficacy of DCB therapy.¹² However, there are no comparable data to the use of CB, and this was also not investigated in our study. In addition, the use of a DCB or a DES had no effect on the TVR rate in our cohort. The ISAR-DESIRE-3 10-year follow-up data could also demonstrate that after PCI for DES-ISR, the primary (composite of cardiac death, AMI, and TVR) and secondary efficacy endpoints (TVR) between DCB and DES were not significantly different.²⁵

It is worth mentioning that, in an optical coherence tomography study, Yin et al showed that an under-expanded stent in an ISR leads to significantly higher rates of cardiac death (3.3%), AMI (9.7%), and TVR (32.4%).¹⁹ This emphasizes the importance of adequate lesion preparation and the use of intravascular imaging. However, the rate of intravascular imaging was only 13% in our cohort, which is more than described in a meta-analysis (4%) in the literature in a real-world scenario,²⁶ but not sufficient to adequately work up our TVR rate.

Limitations. One of the main limitations is the study design (non-randomized registry study) and the small number of patients. Therefore, operator-based decisions cannot be controlled due to the study design. Nevertheless, registry data help to generate hypotheses and define endpoints and randomized studies.

Another limitation is that our results may have been influenced by the operators’ decisions. In particular, the steps of lesion preparation before CB or SB dilatation and after stent implantation were the decision of the interventional cardiologist. Furthermore, the decision to use a DCB or DES was also left to the discretion of the operator. The identification of an ISR due to calcified neoatherosclerosis was based solely on angiographic appearances. The decision to use intravascular imaging was left to the operator and not strictly recommended, which resulted in a rate of 13%. Although this is still significantly more than a meta-analysis in which intravascular imaging was used in only 4% of cases in a real-world scenario,²⁶ the number was not sufficient to perform a meaningful analysis here.

Finally, data allowing a comparison of the crossability of the angioplasty balloons used were not systematically collected and therefore not analyzed. Furthermore, the laboratory values (cardiac biomarkers) were not determined at defined time points. This could have led to the fact that not all intrahospital complications were systematically recorded.

Comparative data on CB dilatation and SB dilatation in lesion preparation of an ISR do not exist. Our multivariate analysis shows that other factors, such as calcification and lesion length, influence the angiographic outcome. In addition, there were no differences in the rates of cardiac death, AMI, and TVR between the 2 balloon types. Randomized studies should be conducted based on these initial data.

From the ¹Heart Centre Trier, Department of Cardiology, Barmherzige Brueder Hospital, Trier, Germany; ²Department of Cardiovascular Diseases, German Heart Centre, Technical University Munich, Munich, Germany.

Acknowledgments: The authors thank Sebastian Runge of the Runge Statistik GmbH for statistical analysis and advice.

Disclosures: Dr Leick receives speaker honoraria from AstraZeneca, Boston Scientific, Shockwave Medical, and Abiomed. Dr Rheude receives lecture fees from Abbott, SIS Medical AG, Translumina, Boston Scientific, and AstraZeneca, unrelated to the current work. Dr Cassese receives lecture fees from Abbott, SIS Medical AG, Translumina and AstraZeneca.  Dr Werner receives speaker honoraria and travel grants, and is an adviser to Abiomed, Boston Scientific, and Shockwave Medical. The remaining authors report no financial relationships or conflicts of interest regarding the content herein.

Address for correspondence: Juergen Leick, MD, Heart Centre Trier, Barmherzige Brueder Hospital Trier, Nordallee 1; 54292 Trier, Germany. Email: j.leick@bbtgruppe.de