consistent with those from the crude analysis. Upon multivariate analysis, the most powerful predictors of TLF were chronic kidney disease in non-DM (OR 4.24, 95% CI: 2.07-8.70, P<0.001) and stent type in DM (OR 2.76, P=0.005). Conclusions: This large matched-cohort study demonstrates that PF-AES has better safety and efficacy profile than BD-BES in patients with DM. Background: In the bench testing of cross-over stenting in the coronary bifurcation model, the proximal optimization technique

more »

... T) provided more symmetrical stent expansion and less malapposition in the proximal main vessel (MV). However, few systematic clinical studies have been reported.± Purpose: To investigate the efficacy of the POT on the cross-over stenting followed by side branch dilation (SBD) in a prospective multicenter clinical study under the guidance of optical coherence tomography (OCT). Methods: In the 3-D OCT Bifurcation Registry from 10 Japanese institutes, a total of 134 bifurcation lesions in 133 patients treated with crossover stenting followed by SBD were divided into POT (n=52) and non-POT groups (n=82). The OCT observation was performed before and after the intervention as well as after the guide wire recrossing (GWR). We investigated incomplete stent apposition, stent eccentricity index defined as the ratio of minimal to maximal diameter of the stent area, stent expansion ratio of proximal MV to distal MV reference, and incidence of GWR into the optimal distal cell located in the side branch (SB) ostium. The effects of timing of the POT, pre-SBD (n=26), final (n=12), and both performed (double; n=13), were also investigated. Results: There were no significant differences in patient or lesion characteristics between POT and non-POT groups except for lower frequency of diabetes mellitus in the POT group (26% vs. 49%, p=0.02). Similar stent (mean 3.0mm / 23mm) and SB balloon (2.5mm) were used in the groups, and mean size of the POT balloon was 3.5±0.6mm. Kissing balloon inflation was also similarly performed in 89%. Although the POT required more contrast medium (174±56 ml vs. 156±50 ml, p=0.05) and operation time (112±37 min vs. 90±31 min, p=0.001), there were no significant differences between the groups in success rate of GWR into the optimal cell (72% vs. 65%), incidence of the type without any link-connection in the SB ostium (51% vs. 49%), incomplete stent apposition at the bifurcation (13±11% vs. 10±9%) or stent expansion ratio (1.3±0.4 vs. 1.3±0.3). However, stent area of more than 5.7mm 2 in the proximal MV was a cut-off value for predicting the optimal GWR in the POT group (AUC 0.71, 95% CI 0.48-0.94, Figure A) , while there were no significant predictive factors in the non-POT group. Only double-POT provided greater stent eccentricity index in the proximal MV, while either pre-SBD-POT or final-POT did not ( Figure B ). Conclusion: Widely stent expansion in the proximal MV induce by the POT increased a possibility of the optimal GWR and symmetric stent expansion was provided by the double-POT; however, the POT did not reduce stent malapposition in the present clinical study. Optimal SBD and device selection under the OCT-guidance might neutralize the efficacy of the POT. Background: Bifurcation is an important lesion subset encountered in daily practice of interventional cardiology. Recent studies have shown that the clinical outcome of the 2-stent technique is not significantly worse than those of the 1-stent technique in contemporary practice using second generation coronary stents. Purpose: Thick-strut, biodegradable-polymer biolimus A9-eluting stents (BP-BESs) are newer but concerned that long-term clinical outcomes may be worse, especially in complex lesion subsets such as bifurcation, compared to other thinstrut drug eluting stents using durable but stable polymers. We used a large-scale multicenter stent registries in Korea to verify the hypothesis that BP-BESs were not worse than durable polymer drug-eluting stents (DP-DESs), regardless of the 1-stent or 2-stent technique. Methods: Patients with BP-BESs were recruited from the prospective Biolimus-Korea-3000 registry and DP-DESs were obtained from four other prospective registries (everolimus-eluting stents from the HOST-Excellent and Excellent-Prime registries, zotarolimus-eluting stents from the HOST-Resolinte and Resolute-Korea registries). From January 2004 to November 2014, a total of 3007 (for BP-BESs) and 10165 (for DP-DESs) patients were consecutively enrolled for these 5 large-scale registries with an unrestricted, open-label, multicenter, prospective manner. Of the patients who were followed up to 3 years, only those with bifurcation lesions were included in the analysis. For more practical analysis, propensity score matching was performed to reflect the significant differences on the lesional and procedural characteristics. Clinically-driven target lesion revascularization (CD-TLR) was chosen for the primary efficacy outcome. Results: A total of 787 lesions were followed up to 3 years for BP-BESs, and 2474 patients were followed for DP-DESs. 696 pairs of lesions for each cohort were collected after the matching process. Dedicated 9-or 13-month angiographic followup was performed in only a small number of patients in both groups. (8.9% for both cohorts, p=1.000). Of these matched lesions, 3-year CD TLR rates were not higher than expected (5.5% for BP-BESs and 4.3% for DP-DESs). These rates were comparable to each other even after adjusting the 1-or 2-stenting technique (HR for BP-BESs 1.271, 95% CI 0.787-2.053, p=0.327), and there was no significant interaction between stenting technique and stent type (p for interac-Downloaded from https://academic.oup.com/eurheartj/article-abstract/38/suppl_1/ehx504.P3314/4089816 by guest on 30 July 2018