**5. PCI versus CABG in SIHD**

The introduction of 3-hydroxyl-3-methyl-glutaryl-coenzyme A (HMG-CoA) reductase inhibitors otherwise known as the "statins," has significantly improved the care of patients with coronary artery disease [34, 35]. Simvastatin 40mg orally daily compared to placebo in patients with known vascular disease was shown in the Heart Protection Study (HPS) to *reduce all cause mortality with an ARR of 1.8 percent* over a five year period [34]. This was paralleled

ACEi's have also been established to have a significant benefit towards long-term cardiovas‐ cular outcomes. It was shown in the Heart Outcomes Prevention Evaluation Study that in highrisk patients (vascular disease or diabetes plus one other risk factor) Ramipril compared with Placebo provides a *relative risk reduction* in myocardial infarction, stroke and death from cardiovascular causes of *21 percent* and *ARR of 3.8 percent* [36] over the mean follow up period

Concurrently, there was the advent and evolution of percutaneous approaches to revascula‐ rization. The first generation of angioplasty that truly adopted popular practice involved serial balloon inflations at the site of stenoses restore normal flow dynamics down the conducting epicardial vessels [6]. This approach, although promising was limited by a high rate of restenoses as a result of localized vascular recoil and epithelial hyperplasia [20, 37]. The bare metal stents (BMS) were created as a scaffolding technique that limited the degree of recoil but still faced significant re-stenosis rates due to mediated by injury to the medial layer, increased inflammation resulting from stent strut penetration into the lipid core and ultimately neoin‐

Pacitaxel and Sirolimus DES were developed in the next phase to overcome the challenge of restenosis requiring repeat intervention [20]. Although there was no difference in mortality or rates of myocardial infarction seen with DES compared with BMS, there were considerable

Early enthusiasm for the use of drug eluting stents was curbed by a significantly higher rate of stent thrombosis, particularly in the face of an initially shorter duration (6 months) of dual antiplatelet therapy (DAPT) [22, 42, 43]. Currently the American College of Cardiology/ American Heart Association (ACC/AHA) recommends at least *12 months of DAPT* in patients receiving DES for non-ACS indication and *12 months of DAPT* for ACS indication regardless

The most recent development in stent technology has been the introduction of secondgeneration (everolimus and zotarolimus) drug eluting stents. A large Swedish registry observational study containing 94, 384 patients demonstrated the advantage of the secondgeneration stents over its predecessors (first generation DES and BMS) both in terms of *restenosis* and *definite stent thrombosis* [44]. The second generation DES in this study was shown to have lower risk of restenosis compared with both BMS and the first generation DES with **Hazard Ratios (HR**) of **[0.29, 95% confidence interval (CI) 0.25-0.33]** and **[0.62, CI: 0.53 -0.72]** respectively [44]. The Cobalt Chromium Everolimus eluting stents (CoCr EES) have shown the most promise in reducing stent thrombosis. In a recent network meta-analysis of 50844 patients, the CoCr EES was shown to have a lower rate of 1 year definite stent thrombosis compared to both paclitaxel DES and sirolimus DES with odds ratio (OR) of **[0.41 95% CI 0.24-0.70]** and **[0.28 95% CI 0.16-0.48]** respectively [24]. The CoCr EES also had a lower rate of definite stent thrombosis compared with BMS at 1 year and 2 years with an odds ratio [OR] of

reductions in restenosis rates with an estimated RRR of 0.44 [21, 22, 39-41].

of the stent type (BMS or DES) *(Class I recommendation, Level B evidence)* [14].

of 5 years. There was also a *reduction in all cause death with an ARR of 1.8 percent* [36].

with a **reduction of coronary death with an ARR 1.2%** [34].

timal growth [38].

322 Artery Bypass

There have been numerous randomized studies comparing PCI with surgical revasculariza‐ tion in MVD. A recent systematic review including 10 major trials over and the individual data from over 7800 patients found similar mortality in patients treated with CABG (15%) compared to patients treated with angioplasty (16%) over median survival of 5.9 years (p=0.12) [45]. There was, however a significantly lower rate of death or repeat revascularization in those treated with CABG (9.9%) compared with those treated with PCI (24.5%) [45]. This suggests that in fact the major benefit seen in CABG over PCI in this comparison is a lower need for repeat revascularization and is paralleled by a lower incidence of angina in the CABG group (14%) compared to the PCI group (26%) at one year (p<0.0001) [45]. The major caveat to this data is that stenting (which is known to reduce restenosis rates) only represented 37 percent of the total angioplasty group [45].

Subgroup analysis revealed that patients with diabetes have overall *better survival* when treated with surgical revascularization than when treated with angioplasty **with an ARR of 7.7% over 5 years** [45]. More definitive data supporting the use of CABG in patients with diabetes will be presented in *Section 6.2.* Interestingly, there was also a graded age interaction that was significant (p=0.002) [45]. For patients younger than 55 years of age, mortality was lower with PCI (8%) than CABG (10%). In patients between ages 55-64, PCI and CABG had similar mortality rates at 15 and 14 percent respectively [45]. And for patients older than 65 years of age, CABG had a lower overall mortality (20%) compared with PCI (24%) [45]. Prior to this study, this interaction had not been previously reported and we can only speculate whether the effect is a true function of age.

Other subgroups did not prove to contribute any significant interaction to the overall treatment effect [45]. Six of the trials included had POBA as the main mode of PCI whereas four trials used BMS; neither of these groups had significantly different survival rates when compared with surgery [45]. There was no overall interaction contributed by the presence/absence of proximal LAD disease, 3VD, abnormal LV function, previous MI or unstable symptoms [45].
