Background Second-generation drug-eluting stents (DES) are said to be more safe and effective. Our aim was to assess clinical outcomes following first- vs second-generation DES implantation in patients undergoing unprotected left main (ULMCA) percutaneous coronary intervention.

Methods All consecutive patients from our single-center prospective registry treated for ULMCA with DES implantation from January 2005 to November 2010 were analysed. The study endpoint was major adverse cardiac event (MACE) defined as all-cause mortality, target lesion revascularisation (TLR), and target vessel revascularisation (TVR) at clinical follow-up.

Results A total of 179 patients were included: mean age 66.5±12.7 years and 83.2% were male with mean left ventricular ejection fraction 54.6±8.4% and SYNTAX score 23.3±31.2. The median follow-up was 705.5 days (IQR 339.8–1168.0). First-generation DES were used in 53.1% (of which 51.6% were sirolimus and paclitaxel 48.4%) and 46.9% had second-generation DES (85.7% everolimus; 11.9% zotorolimus; 2.4% biolimus). Interestingly, there were more patients with diabetes treated with first-generation (30.5% vs 14.3%; p=0.023). Regarding the procedure, intravascular ultrasound guidance was similar between first and second-generation (respectively 47.4% vs 59.5%; p=0.130). There were more patients in the first-generation group with distal ULMCA disease (82.1% vs 67.9%; p=0.064). At follow-up, there was a significant difference in MACE favouring second-generation (30.5% vs 19.0%; p=0.047), most related to a reduction in the TLR (13.7% vs 4.8%; p=0.026) and TVR (24.2% vs 14.3%; p=0.031). However, there was no difference in all-cause mortality (10.5% vs 7.1%; p=0.138) with a trend for increased cardiovascular mortality in those treated by first-generation (8.4% vs 2.4%; p=0.082). Moreover, there was no difference in definite/probable stent thromboses (5.3% vs 2.4%; p=0.114).

Conclusions Second-generation DES have improved results with regards to MACE at mid-term follow-up, perhaps secondary to patient selection. This needs to be confirmed at longer-term follow-up.