Impact of stent diameter and length on in-stent restenosis after drug eluting stent versus bare metal stent implantation in patients needing large coronary stents

Abstract

Introduction: The advantage of drug eluting stents (DES) compared to bare metal stents (BMS) is small for the the risk of in-stent-restenosis (ISR) in short lesions of large coronary arteries. Therefore, the British National Institutes of Clinical Excellence (NICE) guidelines recommend to use BMS instead of DES in lesions 3mm diameter. However, no dedicated trial has tested this question. We analysed the impact of stent length and stent diameter on ISR in the BASKET-PROVE study population. Methods: In the multicentre randomized BASKET-PROVE study we compared BMS vs. DES regarding target vessel revascularization (TVR) rates in large coronary arteries (needing ≥3mm stents) resulting in significantly lower 2-year TVR rates in the DES group. For this retrospective subgroup analysis two experienced interventionalists blinded to stent type reviewed all angiographic films for the presence of ISR. We used a mixed-effects Cox survival model to analyse the impact of stent length and stent diameter on ISR on a basis, with ISR being defined as luminal narrowing >50% by visual assessment. Results: 2312 patients with a total of 3278 stented segments were included in the analysis (mean age 63.7±10.9 years; 24.4% female). 123 patients had at least one re-angiography during the two years follow-up due to recurrent ischemia. A total of 75 ISR in 63 patients were observed (mean age 61.9±11.1 years). ISR was significantly more frequent in segments with a BMS (5.3%) compared to segments with a DES (0.74%; HR = 6.18, p<0.001). Based on the Cox analysis, the strongest predictor of ISR was the use of a BMS (hazard ratio (HR)) 6.74; p<0.001). There was a decreased risk of ISR per 1mm increase in stent diameter (HR of 0.30, p=0.003). The HR for stent length per segment was 1.02 (p=0.02; CI= 1.0-1.04) indicating a 2% increased risk of ISR per 1mm increase in stent length. Co-factors like renal failure or diabetes mellitus had no significant effect on ISR. The HR for ISR in patients with stent diameter >3mm and stent length <15mm was not significantly different between the DES and the BMS group although the event rate in this subgroup was low (2.0% with BMS, 1.4% with DES, p=NS). Conclusion: The risk of ISR in patients receiving stents in large coronary arteries is low. Nevertheless, even in patients in need of stents ≥3.0mm, the strongest predictor of ISR was the use of a BMS. Other co-factors like renal failure or DM are not associated with ISR in this population. Alltogether, these findings challenge the restriction of NICE to use BMS instead of DES in vessels >3.0mm.