There are several observational studies comparing BMS and DES in LMD. In most, event free survival was improved with DES. In the LE MANS [7] (Left MaiN coronary artery Stenting) registry, performed in Poland, 252 patients were enrolled during 11 years between 1997 and 2008 with 58 % having a non ST segment elevation acute coronary syndrome, diabetics in 26.4 % and a distal location of the stenosis in 56 % [8]. Between 1997 and 2001 only BMS were used. After that, DES was recommended for the LM with a reference diameter ≤3.8 mm. First generation DES [paclitaxel (PES) and sirolimus (SES)] were implanted in 36.2 % of patients. The Euro score surgery risk was 6 ± 2.

Major adverse cardiovascular and cerebral events (MACCE) defined as death, myocardial infarction (MI), TLR, stent thrombosis (ST) or cerebrovascular accident (CVA) stroke occurred in 12 (4.8 %) patients during the 30-day follow-up period. During long-term follow-up (mean 3.8 years, range 1–11 years) MACCE rates were 25.4 % and 13.9 % of patients died. The 5- and 10-year survival rates were 78.1 % and 68.9 % respectively. Despite more favorable baseline characteristics in patients treated with BMS, unmatched analysis showed a significantly lower MACCE rate in DES patients (25.9 % vs. 14.9 %, respectively p = 0.039). This difference was strengthened even further after propensity score matching. DES lowered both mortality and MACCE for distal LMD lesions when compared to BMS. Ejection fraction <50 % was the only independent risk factor influencing long-term survival.

The Italian Society of Invasive Cardiology ran a multicenter retrospective registry [9] in 19 high volumes centers enrolling 1453 consecutive patients who underwent PCI on LMD between January 2002 and December 2006. Four hundred and seventy-nine received either first generation DES (334 patients) or BMS (145 patients) but only for lesions located at the ostium or shaft, without any distal involvement. After propensity score matching was performed, baseline covariates between the two yielded 119 well-matched pairs. At 3-year follow-up, risk-adjusted survival rates were higher in patients treated with DES than in those treated with BMS, although the adjusted 3-year rates of TLR were not significantly lower with DES compared to BMS (P = 0.60).

Our group began to perform left main stenting in the early 1990s soon after PCI with BMS became a default strategy [4]. In 1996, the ERACI II study [10, 11] became the first randomized revascularization trial that included unprotected LMD patients. This included about 5 % of the overall population, 4 % in CABG and 5.8 % in the PCI arms. Moreover, in Argentina in 2013, a multicenter registry was published in patients with unprotected LMD [12]. Two hundred and eighty-one consecutive patients treated between 2002 and 2012 were included. Baseline characteristics included mean age of 67.1 years, diabetes in 18.5 %, a Euro SCORE of 5.5, acute coronary syndromes in 72.9, 25.4 % of them with ST elevation MI (STEMI) and distal LMD stenosis in 49.8 %. 391 stents were implanted, BMS and DES in 49 % and 51 % respectively. During a mean follow-up 3.2 ± 2.7 years, death occurred in 9.6 and 6.0 % had an MI; 14.5 % had death/MI/CVA and 30 % MACCE (Fig. 16.1). Compared to proximal lesions, distal lesions had a significantly worse freedom from death/MI/CVA (81.9 % vs. 89 %, respectively = 0.012) (Fig. 16.2) and MACCE (66.7 % vs. 77.7 %, respectively p = 0.02). Patients receiving a DES had less MACCE than with BMS (p = 0.025). Those treated with BMS also had poorer outcomes with distal lesions (Fig. 16.3). Better event free survival with DES was also observed in patients treated for an acute MI with ostial LAD or LCX involvement. Thus, DES implantation had become the default strategy in left main PCI [13]. Further pooled data from several trials suggested better outcomes with DES compared to BMS. A meta-analysis [14] with 10,342 patients from 44 studies assessed the incidence of death/MI, TVR/TLR and MACCE at 3 years. Mortality (8.8 and 12.7 %, p = 0.01), TVR/TLR (8.0 and 16.4 % with DES and BMS: p = 0.01), and MACCE (21.4 and 31.6 %, p = 0.12) were all lower with DES.

There are few studies that compared different DES designs in LMD. A randomized trial [14] and two observational studies [15, 16] suggested that outcomes were comparable between both first generations DES. However, now we are using second and third generation drug- (limus family mostly, sirolimus, everolimus, zotarolimus and biolimus) eluting stents with new polymers, including biocompatible or biodegradable polymers. In the Florence registry, 390 patients underwent PCI with DES implantation, 224 received a paclitaxel eluting stent (PES) and 166 an everolimus-eluting stent (EES) [17]. At 9 months, coronary restenosis by angiography was 5.2 % in EES and 15.6 % in PES (p = 0.002). Among 166 propensity matched pairs, the rate of MACCE at 2 years was significantly higher with PES (20.4 versus 10.2 %, respectively, p = 0.010).

In the Left main Taxus and left main Xience (LEMAX) non-randomized registry, 173 patients treated with an EES were compared with a historical cohort of 291 patients treated with PES. At 12-month clinical follow-up, the EES was associated with a lower rate of target lesion failure, MACCE and ST compared with PES [18]. The ISAR-LEFT-MAIN 2 study compared the safety and efficacy of the zotarolimus-eluting stent (ZES) versus EES [19]. Patients were randomly assigned to receive either a ZES (n = 324) or an EES (n = 326). The primary endpoint was the combined incidence of death, MI and TLR at 1 year. Secondary endpoints were definite or probable ST at 1 year and angiographic restenosis based on analysis of the left main coronary artery area at follow-up angiography. At 1 year, the cumulative incidence of the primary endpoint was 17.5 % in the ZES group and 14.3 % in the EES group (p = 0.25). Three patients in the ZES group (0.9 %) and 2 patients in the EES group (0.6 %) experienced definite or probable ST (p < 0.99). All-cause mortality at 1 year was equal in the 2 groups (5.6 %). Angiographic restenosis occurred in 21.5 % in the ZES group and 16.8 % in the EES group (p = 0.24). The results suggested that the use of second-generation EES and ZES had comparable outcomes to those noted with the use of first-generation DES, and both stent types appeared to afford similar results at 1-year follow-up.

There are several non-randomized studies and registries (Table 16.1) comparing DES to CABG for patients with LMD [21–23]. The DELTA multicenter registry [20] included consecutive “all comers” with LMD treated either with PCI and a 1st generation DES or CABG between April 2002 and April 2006. Patients treated in 14 centers were retrospectively analyzed in this worldwide registry. A propensity score analysis was performed to adjust for baseline differences in the overall cohort. In total 2775 patients were included: 1874 were treated with PCI and 901 with CABG. At 1295 days (range: 928–1713), there were no differences, in the adjusted analysis, in the primary composite endpoint of death, cerebrovascular accidents, AMI, or the composite endpoint of death and AMI . An advantage of CABG over PCI was observed in the composite secondary endpoint of MACCE (p = 0.0001), driven exclusively by the higher incidence of TVR with PCI.