Impact of Diabetes on 10-Year Outcomes of Patients With Multivessel Coronary Artery Disease in the Medicine, Angioplasty, or Surgery Study II (MASS II) Trial

Eduardo Gomes Lima, MD; Whady Hueb, MD, PhD; Rosa Maria Rahmi Garcia,MD, PhD; Alexandre Costa Pereira,MD, PhD; Paulo Rogério Soares, MD, PhD; Desiderio Favarato, MD, PhD; Cibele Larrosa Garzillo, MD, PhD; Ricardo D'Oliveira Vieira, MD; Paulo Cury Rezende, MD; Myrthes Takiuti, RN, PhD; Priscyla Girardi, RN, PhD; Alexandre Ciappina Hueb, MD, PhD; José A. F. Ramires, MD, PhD; Roberto Kalil Filho,MD, PhD


Am Heart J. 2013;166(2):250-257. 

In This Article


In a previous publication[17] with a similar population but in a 5-year follow-up, we found that the initial therapeutic approach, medical treatment, angioplasty, or surgery, did not change the mortality trend over 5 years for non-diabetic subjects with stable multivessel coronary disease. Besides, the treatment modality did not influence the outcomes during the first year in diabetic subjects. However, from the first year and afterward up to 5-year follow-up, diabetic subjects undergoing treatment with invasive strategies (angioplasty or surgery) had significantly improved mortality rates in comparison with patients randomized to a more conservative medical strategy.

The major finding of this present study is that in a 10-year follow-up we found a statistically significant difference in overall and cardiac mortality rates favoring CABG over MT among diabetic patients with stable multivessel CAD. In addition, PCI superiority over MT demonstrated in a 5-year follow-up was not seen in a 10-year follow-up.

Various factors have been proposed to explain these results. First, we have to consider the extent of CAD, concerning the angiographic complexity of lesions in the diabetic population. A substudy of diabetic patients in the Syntax trial[18] showed lower mortality rates favoring CABG over PCI using DES in the specific subgroup of patient with Syntax scores >33. It is known that CABG is superior to PCI or MT in more advanced CAD subgroups. Once diabetic patients have more complex lesions, a greater plaque burden, and more 3-vessel disease, we could consider the possibility of the benefit of CABG in this population to be due to more CAD being angiographically complex, and not necessarily to diabetes status.[18,19]

Many studies also have shown more adverse events related to angioplasty in a diabetic population,[20] even in recent trials using DES.[18,21] In fact, diabetes is recognized as a risk factor for stent restenosis and thrombosis.[22] Association between complications related to angioplasty and higher rates of mortality have been described among diabetic patients with chronic CAD.[20,23]

Noteworthy is the recently published result of the 10-year follow-up of the MASS II[24] trial, regarding the total population of 611 patients. Compared with CABG, MT was associated with a significantly higher incidence of subsequent myocardial infarction, a higher rate of additional revascularization, a higher incidence of cardiac death, and consequently a 2.29-fold increased risk of combined events. PCI was associated with an increased need for further revascularization, a higher incidence of myocardial infarction, and a 1.46-fold increased risk of combined events compared with CABG. We have to consider this fact taking into account the impact of diabetes in the progression of atherosclerosis in this subgroup of patients in a long-term follow-up.

The apparent positive effect of CABG among population despite diabetes status could reside in the completeness of revascularization and the use of LIMA in this subset of patients. We found significantly fewer stents per patient than grafts per patient in PCI and CABG groups, respectively, and consequently more incomplete revascularization in the PCI compared to the CABG group. Association between incomplete revascularization and higher rates of mortality has been described in very long-term follow-up studies.[25] Use of LIMA has been associated with better long-term survival in recent trials when compared to CABG without internal thoracic artery grafts,[26] or even compared with DES.[27]

Recent studies have shown conflicting findings in terms of survival rates. BARI 2D showed no difference in survival rates in diabetic patients with multivessel CAD comparing interventional therapies versus medical-therapy (88.3% vs 87.8%, respectively). However, in the CABG stratum, the rate of major cardiovascular events was significantly lower in the revascularization group. It is important to point out that the trial was not designed to compare CABG with PCI. Indeed, randomization for intervention or exclusively medical treatment was performed after the clinical decision for an interventional procedure. Then, patients whose choice of interventional procedure was CABG had more severe CAD than those in chosen for PCI.

The ARTS trial 5-year follow-up substudy was recently published comparing outcomes for 3 interventional strategies (CABG, PCI using BMS, and DES) in 367 diabetic patients with multivessel CAD included in ARTS I and II.[21] There was no significant difference in mortality between the 3 groups. There was, however, a statistically significant difference in the myocardial infarction rate between BMS and CABG arms, favoring the last one. However, we have to consider the larger proportion of 3-vessel CAD disease among our population (59% in MT, 68% in PCI, 54% in CABG arms) than in patients in the ARTS trial (31% in BMS group, 50% in DES, 35% in CABG group). Additionally, the longer follow-up of this present study must be highlighted.

CARDia[28] compared CABG and PCI (performed with DES and BMS) in 510 diabetic patients and multivessel CAD. Results of 1-year follow-up did not find a difference in mortality rates between the groups. Once more, we have to comment on the importance of the duration of follow-up as an important factor to understand CARDia results.

Most recently, FREEDOM trial[11] compared CABG versus PCI using DES in 1900 diabetic patients with multivessel CAD followed for a minimum of 2 years. CABG was superior to PCI with reduced rates of death and myocardial infarction, but higher rates of stroke. The superiority of CABG over PCI must be interpreted considering the population of this study, composed mainly by patients with 3-vessel disease. Once again, it is important to mention that in FREEDOM trial there was not exclusive MT arm.

Limitations of the study should be mentioned. This is a post-hoc analysis of MASS II trial patients with stable angina and preserved LVEF. Thus, findings must be interpreted in the pre-DES era in a long-term follow-up. MASS II trial was not addressed to detect differences in mortality in treatment regimens among diabetic patients, so the present work must be interpreted in this perspective. These results are not generalizable to current diabetic patients who receive a different level of care in terms of global risk reduction and coronary artery management. Patients underwent treatment regimens as they were 10 years ago, and treatments have evolved over time. Balloons and BMS were used instead of DES in the PCI arm. Only On-pump CABG was performed. Patients were treated to reach therapeutic goals, mainly regarding lipid levels and blood glucose management, according to guidelines from 10 years ago to now. On the other hand, all treatment options had these limitations, which can dilute possible biases. It is important to highlight that the medical treatment regimen dispensed to all groups in relation to medical usage and therapeutic goals in follow-up are comparable to those in some contemporary trials.[10,20] More accurate scores to assess angiographic complexity of lesions, like the SYNTAX score, and coronary artery jeopardy scores were not performed in this population. This has to be considered in the interpretation of the results as well in the applicability of them in the current era. Besides, interpretation of the results has to take into account the completeness of revascularization in patients who underwent interventional procedures.