Performance of Thin-Strut Stents in Non-Left Main Bifurcation Coronary Lesions: A RAIN Subanalysis

Objectives. This study assesses the safety and efficacy of thin-strut stents in non-left main (non-LM) bifurcation coronary lesions. Background. Thinner struts of recent drug-eluting stent (DES) devices are associated with improved outcomes, but data about their performance in challenging scenarios are scant. Methods. RAIN was a retrospective multicenter registry enrolling patients with coronary bifurcation lesions or left main (LM) disease treated with thin-strut DESs. Target-lesion revascularization (TLR) was the primary endpoint, while major adverse clinical event (MACE) rate, a composite of all-cause death, myocardial infarction (MI), target-vessel revascularization (TVR), TLR, and stent thrombosis (ST), and its single components were the secondary endpoints. Multivariable analysis was performed to identify predictors of TLR. Outcome incidences according to stenting strategy (provisional vs 2-stent technique), use of final kissing balloon (FKB), and intravascular ultrasound/optical coherence tomography optimization were further investigated in prespecified subanalyses. Results. A total of 1803 patients (59% acute coronary syndrome, 41% stable coronary artery disease) with non-LM bifurcations were enrolled. After a median follow-up of 12 months, TLR incidence was 2.5% (2.2% for provisional stenting and 3.5% for 2-stent technique). MACE rate was 9.4% (all-cause death, 4.1%; MI, 3.2%; TVR, 3.7%; definite ST, 1.1%). After multivariable adjustment, postdilation (hazard ratio [HR], 0.32; 95% confidence interval [CI], 0.15-0.71; P<.01) and provisional stenting (HR, 0.62; 95% CI, 0.55-0.89; P=.03) were associated with lower TLR rates. FKB was associated with a lower incidence of TLR in the 2-stent subgroup (P=.03). Intracoronary imaging had no significant impact on the primary endpoint. Conclusions. Thin-strut DES options represent an effective choice in bifurcation lesions. Postdilation and provisional stenting are associated with a reduced risk of TLR. FKB should be recommended in 2-stent techniques.

J INVASIVE CARDIOL 2021;33(11):E890-E899.

Key words: coronary lesions, final kissing balloon, non-LM bifurcation, provisional stenting, thin-strut DES

Percutaneous coronary intervention (PCI) with drug-eluting stent (DES) implantation is regarded as a standard of care for appropriately selected patients suffering from coronary artery disease (CAD).^1-3 However, in some scenarios, such as coronary bifurcation lesions, PCI is associated with lower procedural success rates and a higher likelihood of in-stent restenosis.^4,5 Although the advent of DES options reduced the global need for repeated revascularization,^6-8 an increasing number of randomized trials have demonstrated that stent features are significant factors of restenosis.^2,9,10 Reduction of strut thickness is among the innovations in DES design conceived to reduce restenosis. Thinner stent struts are associated with a reduction in the shear stress of coronary walls, leading to a reduction of inflammation and of activation of atherosclerosis process,¹¹ translating clinically into a lower rate of revascularization and stent thrombosis (ST).^12,13 Although the efficacy of 3 different devices (Synergy, Orsiro, and Resolute) with thin struts (<100 µm) was tested in the BIORESORT trial with satisfactory results,¹⁴ limited data are available regarding the effectiveness of these devices in coronary bifurcation lesions. In the BIORESORT trial, among an all-comers population, only 35% of treated patients were treated for bifurcation lesions. Moreover, BIORESORT focused on 3 thin-strut DES devices, whereas more devices with similar features are currently available. Therefore, the goal of the present analysis of the RAIN (veRy thin stents for patients with left mAIn or bifurcatioN in real life) registry was to evaluate the incidence of target-lesion revascularization (TLR) in real-world patients with coronary bifurcation lesions, as well as the potential benefit of PCI strategies and optimization techniques to reduce subsequent adverse events.

RAIN (ClinicalTrials.gov NCT03544294) is a retrospective, multicenter study (Appendix 1) including patients from June 2015 to January 2017.

Inclusion criteria. All consecutive patients presenting with a critical coronary lesion involving an unprotected left main (ULM) or a coronary bifurcation and undergoing PCI with a thin-strut DES were included.

Stent types. The devices adopted in the RAIN registry are as follows: (1) the Promus Element stent (Boston Scientific), a platinum-chromium stent coated with a permanent polymer loading everolimus with strut thickness of 81 µm for diameters from 2.25 to 3.5 mm and of 86 µm for a diameter of 4.0 mm; (2) the Ultimaster stent (Terumo Corporation), a cobalt-chromium stent coated with a biodegradable polymer abluminal coating loading sirolimus with strut thickness of 80 µm; (3) the Synergy stent (Boston Scientific), a platinum-chromium stent coated with a biodegradable polymer loading everolimus with strut thickness of 74 µm for diameters in the range 2.25-2.75 mm, 79 µm for diameters in the range 3.00-3.50 mm, and 81 µm for diameters measuring 4.0 mm; (4) the Xience Alpine stent (Abbott Cardiovascular), a cobalt-chromium stent coated with a permanent polymer loading everolimus with a strut thickness of 80 µm; (5) the Resolute Onyx stent (Medtronic), a platinum-chromium stent coated with a biodegradable polymer loading zotarolimus with a strut thickness of 74 µm for diameters ≤2.5 mm, 79 µm for diameters in the range 3.0-3.50 mm, and 81 µm for diameters measuring 4.0 mm; and (6) the Biomatrix Alpha stent (Biosensors), a cobalt-chromium stent coated with a biodegradable polymer loading biolimus with a strut thickness of 84 µm.

Definitions. Critical LM stenosis was defined according to guidelines as visual stenosis of >50% at coronary angiography, or critical area at intravascular ultrasound (IVUS) evaluation, or positive values of fractional flow reserve, according to evaluation by the physician at a single institution. Complex bifurcations were defined according to the following criteria (major + any 2 minor), with (1) major defined as  distal LM bifurcation sidebranch (SB) diameter stenosis (DS) >70% and SB lesion length ≥ 10 mm; and (2) minor, defined as moderate to severe calcification; multiple lesions; bifurcation angle <45° (evaluated by angiography); main vessel reference vessel diameter (RVD) >2.5 mm; thrombus-containing lesions; or main vessel (MV) lesion length ≥25 mm.

Baseline and procedural data. Cardiovascular risk factors, clinical presentation, angiographic features, and procedural details, such as the use of IVUS or optical coherence tomography (OCT), were collected in an anonymous fashion on a dedicated electronic dataset. The study was approved by the local institutional review boards and the participants provided general written informed consent upon admission for data collection and future publication in anonymous studies. Postdilation, final kissing balloon (FKB), use of imaging, and the choice of stenting technique (provisional vs 2-stent technique defined “as treated”) were left to the treating physician’s discretion.

After the index procedure, dual-antiplatelet therapy (DAPT) was indicated for all patients according to European Society of Cardiology guideline recommendations.¹⁵ Specifically, in the case of an acute coronary syndrome, DAPT consisted of acetylsalicylic acid and ticagrelor or prasugrel for 12 months. In the case of stable CAD, DAPT with acetylsalicylic acid and clopidogrel was indicated for at least 6 months. In patients with an indication for oral anticoagulation, triple therapy with either warfarin or a new direct oral anticoagulant in combination with clopidogrel (6 or 12 months) and acetylsalicylic acid (1 month) was recommended. Follow-up was performed through dedicated clinical assessment (preferred), telephonic follow-up, or formal query to primary care physicians.

Endpoints. TLR rate, defined as critical restenosis (angiographic or with imaging or with functional evaluation according to operator’s choice) was the primary endpoint. MACE, a composite endpoint of all-cause death; myocardial infarction (MI) excluding periprocedural MI; TLR; and definite ST according to the definition by the Academic Research Consortium;¹⁶ and its single components, along with target-vessel revascularization, were the secondary endpoints. Outcomes were appraised according to stenting strategy (provisional vs 2-stent), use of FKB, and use of imaging.

Data collection. Data were derived from electronic records at each center on prespecified forms and recorded online in our institutional research website, Cardiogroup.org (http://www.cardiogroup.org/RAIN/).

Statistical analysis. Categorical variables are reported as count and percentage, and continuous variables are reported as mean ± standard deviation or median and interquartile range (IQR). Normality distribution was evaluated by Kolmogorov-Smirnov test. The t-test was used to assess differences between parametric continuous variables, Mann-Whitney U test for non-parametric variables, the Chi-square test for categorical variables, and Fisher’s exact test for 2 x 2 tables. Cox multivariable analysis was performed to assess independent predictors of TLR. According to Peduzzi et al,¹⁷ five variables could be tested in a multivariate model according to the incidence of the main endpoint in the present study without a significant risk of bias (FKB, stenting strategy, use of imaging, diameter of lesions, postdilation). Survival analyses were performed for provisional vs 2-stent strategy, FKB, and use of imaging. A two-sided P-value <.05 was considered statistically significant; all analyses were performed with SPSS, version 21.0 (IBM).

A total of 1803 patients with coronary bifurcation lesions treated with thin-strut DES implantation were enrolled and considered for the present analysis. Most patients (1063; 59%)were admitted for acute coronary syndrome (ACS), with the remainder admitted with stable angina (19%), positive stress test (13%), or planned angiographic follow-up (7%). Demographic and baseline clinical features are displayed in Table 1. Patients were mostly middle aged (mean age, 68 ± 11 years), males (76%), and with multiple risk factors for ischemic heart disease.

Table 2 and Table 3 summarize the coronary lesion characteristics and procedural data, respectively. Briefly, most patients (67%) were treated for bifurcations involving the left anterior descending coronary artery and diagonal branches. Overall, 642 patients (36%) had diffuse coronary artery disease and 165 (9%) had severely calcified lesions. The mean diameter of the main vessel treated was 2.9 ± 0.45 mm, with a mean length of 24.5 ± 9 mm. Bifurcation type according to Medina¹⁸ classification was available for 823 patients (45%). In most cases, PCI was undertaken for true coronary bifurcations, with Medina 1,1,1 bifurcations being the frequent setting.

Provisional stenting was adopted in 1527 patients (85%), whereas a 2-stent technique was performed in 14% of cases. Postdilation was performed in more than half of cases (960; 53%) and FKB was used in 607 patients (37%). PCI was supported by intracoronary imaging in 524 patients (505 with IVUS [28%] and 19 with OCT [1%]). All patients were discharged on DAPT; 1053 patients (58%) were prescribed clopidogrel, while 647 received a potent P2Y₁₂ inhibitor (153 prasugrel [8%] and 494 ticagrelor [27%]), with a mean DAPT duration of 11 ± 3 months.

Clinical outcomes. Follow-up was available for 1685 patients (93%). The study flow chart is shown in Figure 1. The median follow-up was 12 months (IQR, 7-18), with a global TLR rate of 2.5% (42 patients). Patients with TLR had a mean age of 66 ± 10 years and no significant differences among baseline features were identified between this subpopulation and patients without TLR, except for a higher incidence of previous coronary artery bypass grafting in the former. Mean vessel diameter of the main branch was significantly lower in patients with TLR (2.9 ± 0.44 mm vs 2.8 ± 0.40; P=.01). On the other hand, patients experiencing TLR had longer lesions on the sidebranch (16 ± 6 mm vs 20 ± 6 mm; P<.001).

Patients with TLR were less likely to receive postdilation (40% vs 70%; P<.001), whereas rotational atherectomy was more commonly used in the TLR subgroup vs the non-TLR subgroup (9% vs 2%, respectively; P<.01). Figure 2 shows patient outcomes according to procedural strategy. TLR rates of 2.2% and 3.5% were observed in the provisional stenting group and 2-stent technique group, respectively. The rate of TLR among patients who had PCI supported by intravascular imaging was equal to the group of patients who had no imaging (2.5% for both subgroups). Similar rates of TLR were observed among patients treated with FKB (2.3%) and without FKB (2.7%). Figure 3 and Figure 4 show vessels affected by TLR according to the use of intravascular imaging and FKB and stratified according to stent    strategy. Figure 5 shows freedom from TLR at follow-up for FKB, use of imaging, and stenting strategy. FKB was associated with a significantly lower incidence of TLR among patients treated with a 2-stent strategy, but not in those treated with provisional stenting (P=.03). The use of intracoronary imaging was not associated with improved outcomes regardless of stenting strategy.

The global MACE rate was 9.4% (n = 159 patients). Among these, 70 patients (4.1%) died and 55 (3.2%) experienced an MI during follow up. Sixty-two patients (3.7%) underwent TVR. ST occurred in 22 patients (1.3%) during follow-up (19 definite, 1 probable, and 2 possible STs). Detailed information regarding MACE are reported in Supplemental Table S1, Supplemental Table S2, and Supplemental Table S3. Patients with MACE were significantly older, had a higher burden of cardiovascular risk factors, and had a lower ejection fraction compared with the remaining population. Among procedural strategies, radial access and postdilation were less frequently used in patients experiencing MACE at follow-up.

Predictors of TLR. At univariate analysis, admission diagnosis (ACS vs stable CAD; P=.20) and stenting strategy (provisional vs 2-stent technique; P=.50) were not significantly associated with TLR. Similarly, the proportions of patients treated with FKB and with intracoronary imaging support were not significantly lower among the TLR group. On the other hand, postdilation of the treated vessel was associated with the primary endpoint and a significant correlation was observed between the mean diameter of the main branch, the lesion length of the sidebranch, and the incidence of TLR, as stated above. After controlling in a multivariable model for FKB, stenting strategy, use of intracoronary imaging, diameter of lesions, and postdilation, only postdilation (HR, 0.32; 95% CI, 0.15-0.71; P<.01) and provisional stenting vs 2-stent strategy (HR, 0.62; 95% CI, 0.55-0.89; P=.03) were significantly associated with lower TLR rates in the overall population (Figure 6 and Table S3).

The main findings of the present study can be summarized as follows: (1) The use of DES with thin strut thickness (<100 µm) is characterized by a remarkably good safety and efficacy profile in the treatment of coronary bifurcation lesions in real-world patients. (2) The overall incidence of TLR was 2.5%. In particular, we observed a rate of 2.2% and 3.5% of TLR with provisional and 2-stent strategies, respectively. (3) After multivariate adjustment, the use of postdilation and the choice of provisional stenting were associated with lower TLR rates. (4) Globally, the use of intracoronary imaging and FKB did not result in an improved outcome, although FKB improved TLR rates in patients treated with a 2-stent strategy.

To our knowledge, the present study is the first report about thin-strut DES outcomes in coronary bifurcation lesions in a large real-world population of all-comer patients including 6 different DES platforms. Approximately 15%-20% of PCIs are undertaken for coronary bifurcation lesions.⁴ PCI in this setting has traditionally been associated with lower procedural success rates and worse clinical outcomes as compared with non-bifurcation lesions. The effectiveness of DES compared with bare-metal stent (BMS) implantation in reducing the revascularization rate in less-complex coronary lesions has been extended to coronary bifurcations. This was definitively confirmed by a subanalysis of the SCANDSTENT trial,¹⁹ which represents the only randomized, controlled trial available in this area. The SCANDSTENT investigators reported a TLR rate of 5.9% in the subgroup of patients treated with a first-generation sirolimus-eluting stent (Cypher; Cordis) with a strut thickness of 140 um. Similar results were observed in registries that showed a reduction in TLR occurrence in patients treated with DES as compared with a historical BMS control, regardless of whether a 1-stent (TLR, 5.4% vs 36%) or 2-stent strategy (TLR, 8.9% vs 38%) was used.^20,21 As a result, DES became the standard of care for the treatment of coronary bifurcation lesions. Despite several innovations in stent structure over the years and data from the ISAR STEREO trials^9,12 confirming that thinner struts are associated with a net reduction of restenosis, only a few studies evaluated the impact of such improvements in specific and challenging scenarios such as bifurcations. The recent BIORESORT trial¹⁴ showed excellent results of the 3 devices characterized by stent struts ranging from 61-89 µm, reporting rates of 2% for TLR and 5% for MACE in a cohort of 1236 all-comer patients, of whom only 30% had PCI for bifurcation. Of note, the trial was initially conceived to compare devices with permanent or durable polymers. However, all of the stents tested showed excellent clinical outcomes regardless of their polymer composition. Therefore, the SCANDSTENT investigators hypothesized that such remarkable results could be related to the lower strut thickness of the devices, requiring an easier and shorter re-endothelialization process. Results of the present research confirm and extend such results to all patients with coronary bifurcation lesions and to several DES platforms. We found outcome rates comparable to those of the BIORESORT trial (TLR, 2.5% vs 2%; MI, 3.2% vs 2% with everolimus-eluting stent, 3% with zotarolimus-eluting stent, and 3% with sirolimus-eluting stent; definite ST, 1.1% vs 0.4%) with the exception of a higher rate of overall MACE (9.5% vs 5%) mainly driven by a higher all-cause death rate in our study. This difference is likely to be explained by the higher mean age of our population (68 ± 11 years in RAIN vs 63.9 ± 10.8 years in BIORESORT) and a higher prevalence of diabetic patients (25% in RAIN vs 18% in BIORESORT).

A critical issue regarding bifurcation treatment concerns the optimal stenting strategy (provisional vs 2-stent technique). There is an extreme heterogeneity of bifurcation lesions, stenting techniques (ie, culotte, crush, T-stent, Protrusion technique, etc), cohort features, and endpoints explored in the main randomized, controlled trials in this area.^22-24 With this limitation, the current recommended approach is to prefer provisional stenting with the exception of specific scenarios (dissection, reduced flow, or particularly large sidebranch) and bifurcations involving the distal LM, for which a dedicated 2-stent technique (double-kissing double-crush) should be preferred.²⁵ Our results in patients with non-LM bifurcation lesions confirm the previous findings. Indeed, we found provisional stenting to be associated with lower TLR rates, together with stent postdilation.

Our results also confirm currently available evidence about optimization with FKB. Although FKB was not associated with lower TLR in the overall cohort, we found FKB to reduce TLR at follow-up among patients treated with 2-stent technique, but not in those treated with provisional stenting. The issue of whether simultaneous balloon dilations of both main vessel and sidebranch should be performed after stent deployment in the main vessel was first addressed in the Nordic-Baltic Bifurcation Study III.²⁶ In this trial, 477 patients with either stable or unstable disease and a bifurcation lesion were randomly assigned to either FKB or no FKB after stenting of the main vessel. At 6 months, there was no significant difference in the rate of the primary composite endpoint between the FKB and no FKB groups (2.1% vs 2.5%, respectively). This result is also in line with a previous finding of our group referring to the entire RAIN cohort.²⁷

Finally, only 29% of patients underwent an IVUS/OCT-optimized procedure, with no significant benefit with regard to TLR rate at follow-up, even after stratifying the population according to the stenting strategy. The long-term impact of IVUS-guided non-LM bifurcation stenting was first assessed by Kim et al in 1668 patients undergoing DES implantation with or without IVUS guidance.²⁸ IVUS-guided bifurcation stenting significantly reduced 4-year mortality compared with conventional angiography-guided stenting. In addition, IVUS guidance reduced the development of very late ST in patients receiving DES.²⁷ Chen et al analyzed the difference in 1-year outcomes following 2-stent techniques involving implantation of DES for coronary bifurcation lesions between IVUS-guided and angiography-guided groups.²⁹ Late ST at 12-month follow-up was 0% in the IVUS-guided group vs 4.9% in the angiography-guided group (P=.03), resulting in significant differences in ST-segment elevation MI between the 2 groups (2.4% vs 9.8%; P=.03). Of note, in both studies, the incidence of TLR at follow-up did not significantly differ among the subgroups of IVUS- or angiography-guided procedures. The recently published ULTIMATE trial,³⁰ by Zhang et al, showed a net benefit of IVUS-guided DES implantation compared with angiography guidance in an all-comers population (9% of LM lesions, 25% coronary bifurcations) in terms of clinically driven TLR and definite ST (0.9% vs 2.3%; P=.02). Interestingly, this striking superiority observed at lesion-level analysis was not confirmed at patient-level analysis. Moreover, subgroup analysis showed that patients mostly had benefit from IVUS-guided PCI when treated on LM/left anterior descending lesions rather than on other vessels. In view of the low proportion of imaging-guided interventions among our cohort, we emphasize that no definitive conclusions can be drawn from our results and that further studies are warranted to evaluate the relevance of IVUS and OCT in patients treated with thin-strut DES implantation in bifurcation lesions and to definitely assess the impact of such ancillary techniques on clinical outcome.

Study limitations. The scenario of a real-word registry is invariably associated with a potential inclusion bias. However, we believe that the present study may help to extend results from randomized, controlled trials to real-life settings. Because of the retrospective design of this study, a core-lab revision of all angiographic procedures was not available. This limited the availability of some data, such as Medina classification. However, among collected data, most of the patients were treated for a true bifurcation lesion. For the same reason, data about the proximal optimization technique were not systematically collected, and therefore the impact of such a relevant procedural strategy could not be assessed. The use of intracoronary imaging was somewhat limited in this registry (30% of procedures), but since its use was left to the operator’s discretion, its prevalence is likely to reflect its utilization in the daily practice. Finally, a head-to-head comparison with conventional DES implantation was not performed and was bey

The use of thin-strut DES implantation in non-LM coronary bifurcation lesions is characterized by excellent clinical results, particularly with respect to TLR at mid-term follow-up, regardless of stenting strategy (provisional vs 2-stent technique). Postdilation and provisional stenting were associated with lower TLR rates. FKB significantly reduced the occurrence of TLR at follow-up among patients treated with a 2-stent strategy, but not among those treated with provisional strategy.