IJCS | Volume 31, Nº2, March / April 2018

157 –1 –0.5 1 0 0.5 Favours stent Favours CABG Heterogeneity: Chi 2 = 5.99, df = 4 (p = 0.20); I 2 = 33% Test for overall effect: Z = 0.01 (p = 1.00) Total events 177 87 Total (95% CI) 2193 100.0% 2171 0.00 [–0.02, 0.02] EXCEL 2010 71 913 53 903 41.6% 0.02 [–0.00, 0.04] LE MANS 2008 11 52 16 53 2.4% –0.09 [–0.26, 0.08] NOBLE 2016 36 592 33 592 27.1% 0.01 [–0.02, 0.03] PRECOMBAT 2011 17 279 23 275 12.7% –0.02 [–0.07, 0.02] SYNTAX LM 2013 42 357 50 348 16.2% –0.03 [–0.08, 0.02] Study or Subgroup Stent Events Total Weight Risk Difference Risk Difference M-H, Fixed, 95% CI CABG Events Total M-H, Fixed, 95% CI Figure 6 – Studies including the left main coronary artery. Late mortality: stent versus coronary artery bypass grafting. The size of the boxes is proportional to the number of patients in the trial. The bars represent 95% confidence intervals. The diamond represents the synthesis of the results. Abbreviations: CABG: coronary artery bypass grafting; LE MANS: Left Main Coronary Artery Stenting; SYNTAX LM: left main coronary artery subgroup of SYNTAX (Synergy between PCI with Taxus and Cardiac Surgery); PRECOMBAT: Premier of Randomized Comparison of Bypass Surgery versus Angioplasty Using Sirolimus-Eluting Stent in Patients with Left Main Coronary Artery Disease; EXCEL - Coronary Artery Bypass Surgery for Effectiveness of Left Main Revascularization; NOBLE: Nordic-Baltic-British Left Main Revascularization; Boudriot: study by Boudriot et al.: J Am Coll Cardiol. 2011; 57: 538-545 . Graph obtained using the software Review Manager (RevMan), version 5.3. Andrade et al. Stent versus surgery: randomized trials Int J Cardiovasc Sci. 2018;31(2)152-162 Original Article In the evaluation of the results, it is important to highlight the superiority of PCI in the mortality at 30 days. This is in line with a prior systematic review 17 and with the New York registry. 19 The difference is obviously not applicable to patients with lesions of high angiographic complexity, as seen in the analysis of the survival curves of aggregated results from SYNTAX LM and PRECOMBAT. 20 A greater incidence of stroke in the surgical group had already been suggested in prior systematic reviews, 14,17 and in the light of the data presented here, this fact becomes indisputable. It is worth mentioning a reduced incidence of stroke in more recent studies, reflecting a greater care taken by surgeons while manipulating the aorta. The similarity of the mortality results at 1 year is aligned with a prior systematic review, which included studies of the era prior to drug-eluting stents. 17 The difference favoring surgery in regards to late mortality is consistent with another meta-analysis 16 and also with a recently published collaborative study. 21 It should be emphasized that the difference found was due to the large number of patients with diabetes in the studies of the drug- eluting stent era, which disappeared in the aggregated results when these studieswere excluded. Thesedata confirmthose of the collaborative study byHlatki et al. 15 , which demonstrated a lower overall mortality at 5 years with surgery, but no difference among nondiabetic patients.We should emphasize that the study by Hlatki et al. 15 included trials of the balloon era in which two-vessel disease predominated, while in the present review there was a predominance of three-vessel disease and obstruction of the left main coronary artery. In regards to the results of obstruction of the left main coronary artery, it is important to remember that the group of patients with this type of obstruction comprised for a long time a forbidden territory for angioplasty. LEMANS was the first randomized study that attempted to compare stent and surgery in left main coronary artery obstruction, with results similar or even superior to those with PCI. However, this was a small study (105 patients), which has been criticized for not having used grafting of internal thoracic artery in approximately 25% of the cases. After that, emerged the results of the SYNTAX 12 subgroup with left main coronary artery obstruction and of the PRECOMBAT trial and the study by Boudriot et al. 5 , which led to the improvement of the recommendations of PCI in left main artery obstruction. Despite that, the American guidelines only changed the recommendation to IIA in patients with a low SYNTAX score and IIB in patients with intermediate SYNTAX scores. 18 We should emphasize that such recommendations are restricted to patients with a high surgical risk. In the present study, which combined the results of six studies with 4700 patients, the outcomes of PCI with drug - eluting stents were equal or even greater than those with CABG. In light of these evidence and recent results of NOBLE andEXCEL, we believe that the American and Brazilian guidelines 22,23 may be soon modified to improve the classificationof PCIwithdrug‑eluting stents, mainly in left main coronary artery obstruction. In relation to the results in patients with diabetes, it is important to remember that the evidence contrary to PCI in diabetes has its origin in the balloon era, from occasional