Dr. Pedro Nuno Pereira e Freitas
Dr. João Daniel Silvestre Dias Brito
Patients with diabetes mellitus represent 20-30% of all percutaneous coronary interventions performed. Furthermore, they represent a very complex subgroup of patients since their coronary artery disease is more extensive. Over the years, stent technology has been developed and clinical outcomes have improved mainly in non-diabetic patients. In diabetic patients we continue to struggle to offer them the best results with percutaneous coronary intervention. In addition, diabetic patients with multivessel disease still benefit most with coronary artery bypass graft surgery. Nonetheless, the new generation of stents gives us hope of surpassing the surgical technique in the years to come.
The diabetic patient presents a well-known pattern of coronary artery disease (CAD) – extensive atherosclerosis, a greater number of significant stenoses, longer lesions and more diffuse and distal disease . As such, diabetes mellitus (DM) represents a high-risk condition, and diabetic patients with CAD have higher rates of death, non-fatal myocardial infarction (MI), repeat revascularization and stent thrombosis than non-diabetic patients .
Here, we briefly review the historical benefit of first-generation drug-eluting stents (DES) over bare metal stents (BMS) and the most relevant data on second-generation, biodegradable polymers and bioabsorbable DES. We also review the latest findings that can open new questions on the future of percutaneous coronary intervention (PCI) in diabetic patients with multivessel disease.
The development of the first-generation DES (sirolimus and paclitaxel) was a major breakthrough in PCI, especially in patients with DM, since it reduced the incidence of restenosis in comparison to BMS. In fact, a meta-analysis of 35 trials  showed that first-generation DES decreased target lesion revascularization (TLR) when compared with BMS. Nevertheless, no difference in overall mortality, MI or stent thrombosis was shown. Additionally, a meta-analysis of 6 randomized trials comparing sirolimus DES (SES) versus paclitaxel DES (PES) showed that the former had lower TLR but that they were comparable with respect to the risk of all-cause death, MI or stent thrombosis . At the time, a major debate occurred when evidence showed that first-generation DES had a higher rate of late and very late stent thrombosis than BMS, especially in diabetic patients . Consequently, the need for safer stent technology led to the development of the second-generation DES.
There are two second-generation DES: the zotarolimus DES (ZES) and the everolimus DES (EES). Over the years, numerous studies have been conducted to try to understand which stent is the most suitable for PCI in diabetic patients. However, conflicting results made it difficult to interpret the data at the time.
The Endeavor® zotarolimus DES (E-ZES) (Medtronic, Minneapolis, MN, USA) was the first of the second-generation DES to be tested in patients with diabetes. However, some studies revealed that E-ZES was associated with an increased risk of cardiac death, MI, target vessel revascularization (TVR) and target vessel failure (TVF) when compared with the first-generation DES . Consequently, efforts were made to improve stent technology and the Resolute® zotarolimus DES (R-ZES) (Medtronic) was created with a different polymer that extended the release of the drug to 4 months.
A pooled analysis of the international global RESOLUTE program analyzed the pre-specified diabetic cohort of 878 patients who received R-ZES . The 12-month rate of TVF in the diabetic cohort was 7.8%, significantly lower than the performance goal of 14.5% (p<0.001). After 2 years, the cumulative incidence of target lesion failure (TLF) in patients with non-insulin-treated diabetes was comparable to that of patients without diabetes (8.0% vs. 7.1%). On the other hand, in insulin-dependent patients there was a significantly higher TLF rate (13.7%). In the whole population, including complex patients, rates of stent thrombosis were not significantly different between patients with and without diabetes (1.2% vs. 0.8%).
The EES was also extensively tested in this population. A pooled analysis of the SPIRIT II, III, and IV and the COMPARE trials was performed in order to compare the EES with the first-generation PES according to diabetes status. Patients without DM treated with EES had significantly reduced 2-year rates of mortality (1.9% vs. 3.1%; p=0.01), MI (2.5% vs. 5.8%; p<0.0001), stent thrombosis (0.3% vs. 2.4%; p<0.0001), and ischemia-driven TLR (3.6% vs. 6.9%; p<0.0001) when compared with PES. In contrast, there were no significant differences in any measured parameters of safety or efficacy in patients with DM (n=1,869) randomized to EES or PES (mortality 3.9% vs. 2.9%, p=0.27; MI 4.2% vs. 4.9%, p=0.49; stent thrombosis 1.6% vs. 2.0%, p=0.50; and ischemia-driven TLR 5.5% vs. 6.1%, p=0.60). Interestingly, an additional interaction was identified among patients with DM so that the 2-year rate of ischemia-driven TLR was reduced with EES in those not requiring insulin, whereas the opposite trend was observed in insulin-dependent patients. It was hypothesized that these differences could be explained by the different mechanism of action of each drug. While rapamycin analogs (EES and ZES) interfere with cellular mitosis which is tightly regulated by glycosylation-dependent enzymes, paclitaxel interferes with multiple pathways of restenosis, thus explaining why it could be relatively independent of the diabetic state .
The EES was also compared with the SES. The ESSENCE-DIABETES trial  was a randomized study comparing EES (n=149) and SES (n=151) implantation in diabetic patients. The primary endpoint was non-inferiority of angiographic in-segment late loss at 8 months: EES proved to be non-inferior to SES (0.23±0.27 vs. 0.37±0.52 mm; p<0.001 for non-inferiority). At 12 months, ischemia-driven TLR (0.7% vs. 2.6%; p=0.317), death (1.3% vs. 3.3%; p=0.448), and MI (0% vs. 1.3%; p=0.498) were not statistically different between the 2 groups. The SORT-OUT IV substudy on diabetes also revealed the same conclusions . In this study, 390 patients with diabetes were randomized to stent implantation with EES (n=194) or SES (n=196). At 18 months, major adverse cardiac events (MACE) were seen in 10.3% of those treated with EES and in 15.8% of those treated with SES (HR 0.63, 95% CI: 0.36-1.11). Cardiac death occurred in 3.1% vs. 4.6% (HR 0.67, 95% CI: 0.24-1.89), MI in 0.5% vs. 3.6% (HR 0.14, 95% CI: 0.02-1.16), and clinically driven TLR was needed in 3.1% vs. 7.7% (HR 0.40, 95% CI: 0.15-1.02).
There have also been studies comparing the two second-generation DES. In an “all-comer” study with diabetic patients (n=1,855) from 2 prospective registries, the EES (n=1,149) was compared with the R-ZES (n=706). Stent-related outcome was TLF, and patient-oriented composite events were a composite of all-cause mortality, any MI or any revascularization. At 1 year, both TLF (3.7% vs. 3.5%, p=0.899) and patient-oriented composite events (9.1% vs. 10.2%, p=0.416) were similar between EES and R-ZES .
Since most of the aforesaid trials were not adequately powered to detect differences in the clinical outcomes, several meta-analyses and registries were performed to try to establish which stent could be the most suitable for the diabetic patient.
A network meta-analysis of 42 trials published in 2012  analyzed five stent types – BMS, first-generation DES and second-generation DES – in 10,714 patients with diabetes (with 22,844 patient years of follow-up). In this study, all DES were associated with a significant reduction in TVR (37% to 69%) when compared with BMS; however, the efficacy varied with the type of DES. EES was found to be the most efficacious and safe stent and it had the highest probability of being associated with the lowest rate of TVR, MI and any stent thrombosis (87%, 81% and 62%, respectively). Nonetheless, by the time of this analysis there were limited data on R-ZES in the diabetic patient to allow any conclusion on this stent .
Also, the Swedish SCAAR registry  compared the EES with SES or PES for the primary composite endpoint of clinically driven detected restenosis, definite stent thrombosis and all-cause mortality. In 4,751 patients with DM, 8,134 stents were implanted (EES=3,928, PES=2,836, SES=1,370). EES was associated with significantly lower event rates compared with SES (HR 1.99, 95% CI: 1.19-3.08). The same was observed when compared with PES (HR 1.33, 95% CI: 0.93-1.91) but this did not reach statistical significance. These results were mainly driven by a lower incidence of stent thrombosis (SES vs. EES, HR 2.87, 95% CI: 1.08-7.61; PES vs. EES, HR 1.74, 95% CI: 0.82-3.71) and mortality (SES vs. EES, HR 2.02, 95% CI: 1.03-3.98; PES vs. EES, HR 1.69, 95% CI: 1.06-2.72). No significant differences in restenosis rates were observed between EES and SES or PES.
The recent publication of the results of the TUXEDO trial  gave some more insight in this context. In this study, 1,788 patients with DM were randomized to receive PES (889 patients) or EES (899 patients). The primary endpoint was TVF (defined as a composite of cardiac death, target vessel MI or ischemia-driven TVR). At one year, PES did not meet the criterion for non-inferiority to EES with respect to the primary endpoint (5.6% vs. 2.9%; p=0.38 for non-inferiority). In addition, there was a significantly higher 1-year rate in the PES group of spontaneous MI (3.2% vs. 1.2%; p=0.0004), stent thrombosis (2.1% vs. 0.4%; p=0.002), TVR (3.4% vs. 1.2%; p=0.002), and TLR (3.4% vs. 1.2%; p=0.002). Interestingly, in a subgroup analysis, the results among insulin-dependent patients (~40% of the study population) were no different from the results among patients who did not require insulin.
Although second-generation DES seem to provide superior safety and efficacy when compared with first-generation DES, late adverse events remain a problem, presumably because of a persistent inflammatory response in the arterial wall. It has been speculated that the use of stents with biodegradable polymers or without a polymer may avoid late events caused by the chronic inflammation related to its presence .
Interestingly, results were reported in a patient-level pooled analysis  of the three largest randomized trials comparing biodegradable polymer DES (BP-DES) with durable polymer SES – the ISAR-TEST 3, the ISAR-TEST 4 and the LEADERS trials. In this analysis, 1,094 patients with diabetes were included: 657 of these were randomly allocated to treatment with BP-DES while 437 were allocated to treatment with durable polymer SES. At four years, the main finding was that the risk of MACE (25% vs. 26.6%, HR 0.95, 95% CI: 0.74-1.21; p=0.67) and of TLR (15.5% vs. 17.4%, HR 0.89, 95% CI: 0.65-1.22; p=0.47) was equivalent in both groups. Remarkably, definite or probable stent thrombosis occurred less often in those treated with BP-DES vs. SES (2.8% vs. 6.1%, HR 0.52, 95% CI: 0.28-0.96; p=0.04), a difference driven by significantly lower stent thrombosis rates with BP-DES between 1 and 4 years (0.4% vs. 2.8%, HR 0.15, 95% CI: 0.03-0.70; p=0.02). Since this is a post hoc analysis, these results should be considered as hypothesis-generating and cannot be extended to other available durable polymer DES (for example EES). As such, these findings remain to be confirmed in a dedicated randomized controlled trial.
Several polymer-free stents have been created. However, there is only one randomized trial – the RESERVOIR trial  – which aimed to compare the results of an amphilimus-eluting stent (AES) against EES in patients with DM. The AES is a stent manufactured from a cobalt-chromium alloy with an 80 µm strut thickness and an ultrathin (0.3 µm) passive carbon coating. This stent does not have any polymer and the antiproliferative drug (sirolimus) is loaded into reservoirs on the stent’s abluminal surface. In the RESERVOIR trial, 112 patients with DM were randomized to receive AES or EES in a 1:1 ratio. The primary endpoint was the neointimal volume obstruction at 9 months, assessed by optical coherence tomography. In this trial the efficacy of AES proved to be non-inferior to EES (11.97±5.94% for AES and 16.11±18.18% for EES; p for non-inferiority =0.0003). There were no significant differences in any of the pre-specified clinical endpoints (ischemia-driven TLR, TVR, cardiac death and probable or definite stent thrombosis) between the two stents, even though this study was not powered to detect differences in this context. Of note, in this trial only 1 and 2-vessel disease diabetic patients were included.
The bioabsorbable DES (BVS) have the unique feature of complete polymer and scaffold bioresorption after 1 to 2 years. Besides the aforementioned theoretical advantages of polymer-free DES, BVS also have the capability of restoration of vasomotor function, adaptive shear stress, cyclic strain and late luminal enlargement with the disappearance of the scaffold .
Until now, there has been only one pooled analysis  of the ABSORB and the SPIRIT trials whose aims were: 1) to assess the 1-year clinical outcomes of 136 diabetic patients compared with 415 non-diabetic patients treated with the Absorb® BVS (Abbott Vascular, Santa Clara, CA, USA.), and 2) to compare the 1-year clinical outcomes of diabetic patients treated with the Absorb® BVS versus EES using a propensity-score matching analysis. The device-oriented composite endpoint was defined as a composite of cardiac death, target vessel MI or ischemia-driven TLR and the patient-oriented composite endpoint was defined as a composite of all-cause death, all MIs or any repeat revascularization. There were no significant differences between diabetic and non-diabetic patients in 1-year rates of the device-oriented composite endpoint (3.7% vs. 5.1%; p=0.64), the patient-oriented composite endpoint (7.4% vs. 8.2%; p=0.86) or stent thrombosis (0.7% vs. 0.7%; p=1.0). In the propensity score analysis, 102 diabetic patients treated with Absorb® BVS were matched with 172 diabetic patients treated with EES. Again, there were no significant differences between the Absorb® BVS and EES groups in 1-year rates of the device-oriented composite endpoint (3.9% vs. 6.4%; p=0.38), the patient-oriented composite endpoint (7.8% vs. 11.0%; p=0.39) or stent thrombosis (1.0% vs. 1.7%; p=1.0). Nevertheless, these findings must be taken with caution since this was a propensity score analysis with trials that included 1 or 2-vessel disease (patients with 3-vessel disease were excluded) with non-complex lesions. Therefore, the clinical performance of the Absorb® BVS in diabetic patients with complex lesions must be further investigated.
In patients with diabetes mellitus and stable multivessel disease, current guideline recommendations (class I, level A) favor coronary artery bypass graft surgery (CABG) over PCI . However, the studies that showed mortality benefit of CABG over PCI were performed against first-generation DES .
The latest findings of the TUXEDO trial  and of a propensity score study performed in diabetic patients with multivessel disease undergoing PCI with EES versus CABG [13,18] could reignite this discussion. In the latter, 4,048 diabetic patients submitted to PCI with EES were matched with 4,048 diabetic patients submitted to CABG. In the matched cohort, at short-term (30 days) follow-up, EES was associated with a lower risk of death (0.57% vs. 1.11%, HR 0.58, 95% CI: 0.34-0.98; p=0.04) and stroke (0.25% vs. 1.41%, HR 0.14, 95% CI: 0.06-0.30; p<0.0001) but a higher risk of MI (0.44% vs. 0.27%, HR 2.44, 95% CI: 1.13-5.31; p=0.02) when compared with CABG. On the other hand, at long-term follow-up, EES was associated with a similar risk of death (10.50% vs. 10.23%, HR 1.12, 95% CI: 0.96-1.30; p=0.16), with a higher risk of MI (6.42% vs. 4.10%, HR 1.64, 95% CI: 1.32-2.04; p<0.0001), with a lower risk of stroke (2.92% vs. 3.88%, HR 0.76, 95% CI: 0.58-0.99; p=0.04) and with a higher risk of repeat revascularization (21.96% vs. 10.40%, HR 2.42, 95% CI: 2.12-2.76; p<0.0001) when compared with CABG. Nonetheless, patients with severe left main disease were not included in this analysis.
In addition, if the BP-DES, polymer-free DES and bioabsorbable DES prove to be safe and efficacious in the setting of multivessel disease in patients with DM, the gap between PCI and CABG may finally have a chance to be overcome.
The currently available data from numerous studies support the use of EES as the preferred stent in the setting of PCI in diabetic patients. In addition, the newer stent technology that is being developed and tested (BP-DES, polymer-free DES and bioabsorbable DES) shows promise in this setting, especially since these newer stents are starting to show themselves to be safer than second-generation DES.
The question as to whether second-generation DES (in particular EES) or the newer stents (BP-DES, polymer-free DES and bioabsorbable DES) can outperform CABG in the setting of multivessel disease in patients with DM must be studied further with properly conducted randomized controlled trials.
Dr. Pedro Nuno Pereira e Freitas, MD and Dr. João Daniel Silvestre Dias Brito, MD
Division of Cardiology, Hospital de Santa Cruz, Lisbon, Portugal
Author for correspondence:
Dr. Pedro Nuno Pereira e Freitas, MD
Postal address: Avenida Professor Doutor Reinaldo dos Santos, 2790-134 Carnaxide, Portugal
E-mail address: firstname.lastname@example.org
Conflict of interest:
The authors have no conflicts of interest to declare.
Our mission: To reduce the burden of cardiovascular disease
© 2018 European Society of Cardiology. All rights reserved