Percutaneous Coronary Intervention in Patients With Diabetes

Abstract

Patients with diabetes and coronary artery disease represent a challenging and growing subset of the population. Although surgical revascularization is the preferred treatment for patients with diabetes and multivessel coronary artery disease with stable angina, a significant proportion of diabetic patients undergo percutaneous revascularization due to comorbidities, presence of single-vessel disease, or presentation with myocardial infarction. The development of drug-eluting stents has significantly improved the results of percutaneous revascularization among diabetic patients, but a number of challenges remain, including higher rates of restenosis and stent thrombosis among diabetic patients. With current technologies, the outcomes of diabetic patients treated with noninsulin agents have approached that of nondiabetic patients. In comparison, patients with diabetes who require insulin therapy represent a high-risk cohort with increased rates of target vessel failure after coronary revascularization. The development of bioresorbable stents and new drug elution systems may provide additional future benefit among patients with diabetes undergoing percutaneous coronary artery revascularization.

Over 170 million people worldwide have diabetes mellitus (DM), with an increasing proportion of these patients diagnosed in the developing world.¹ Despite medical advances in prevention of coronary artery disease (CAD), the rising prevalence of DM threatens to increase the global burden of cardiovascular disease in the next 2 decades.² Currently, patients with DM compose 25% to 30% of all patients undergoing coronary artery revascularization.³ Up to 60% of patients presenting with acute myocardial infarction may also have previously unrecognized diabetes.^4-6 The increasing recognition that DM is associated with adverse outcomes after coronary revascularization has led to the development of newer stent technologies and pharmacotherapies that may be differentially utilized in diabetic patients.

This article reviews revascularization strategies, the current use of percutaneous coronary intervention (PCI), new developments in stent technology, and the importance of medical therapy after PCI among patients with DM.

Timing of Revascularization and Revascularization Strategy

Patients with DM have more diffuse and complex CAD than nondiabetic patients.⁷ Revascularization of high-risk anatomy including proximal LAD stenosis is also associated with mortality benefit among patients with DM.⁸ Some have therefore advocated performing revascularization in diabetic patients regardless of symptoms. Balanced against these benefits is the observation that patients with DM who undergo revascularization experience less improvement in anginal symptoms compared to nondiabetics.⁹ The BARI 2D trial was the first large randomized trial to enroll exclusively diabetic patients with CAD and stable angina.¹⁰ This study demonstrated that an initial revascularization strategy plus optimal medical therapy yielded outcomes similar to optimal medical therapy with deferral of revascularization. Even with a deferred strategy, however, 38% of patients required revascularization during a 5-year follow-up period.¹¹ Importantly, patients with left main coronary artery stenosis ≥ 50% or uncontrolled angina were excluded from BARI 2D and should undergo prompt revascularization.

For diabetic patients requiring coronary artery revascularization, the choice must often be made between coronary artery bypass grafting (CABG) and PCI with placement of a stent. The recently published FREEDOM trial randomized patients with DM and multivessel CAD to a revascularization strategy of CABG or PCI.¹² During 5-year follow-up, the group randomized to CABG had significantly lower rates of myocardial infarction (6.0% vs 13.9%) and death (10.9% vs 16.3%), but a slightly higher risk of stroke (5.2% vs 2.4%). Similar results were reported in the VA-CARDS trial of patients with DM and multivessel or proximal left main CAD.¹³

The above studies suggest that patients with DM and stable angina due to multivessel CAD should ideally undergo CABG, especially in the presence of hemodynamically significant disease of the left anterior descending coronary artery. In clinical practice, however, many patients with CAD and DM are high risk for surgery due to multiple comorbidities. The patients enrolled in these studies also had a normal left ventricular ejection fraction and minimal or no symptoms of heart failure, whereas many patients with DM and CAD have concomitant depressed ejection fraction and heart failure. In addition, current randomized trials included primarily earlier-generation drug-eluting stents (DES); the improved safety profile of current stent technologies may therefore favor stent placement in certain patient cohorts. The findings of the above trials cannot therefore be extrapolated to all patients with DM or patients with acute coronary syndromes, where PCI has benefit in diabetic patients.^14,15

Limitations of Current Stent Technologies in Patients with Diabetes

Stent technologies have significantly broadened the applicability of angioplasty to patients with complex CAD over the past 2 decades.¹⁶ DES in particular dramatically lowered the rates of restenosis and need for repeat interventions among diabetic patients.¹⁷ Despite these improvements, patients with DM remain more likely to develop restenosis and stent thrombosis than patients without DM. These 2 modes of recurrent cardiovascular events remain important limitations of current stent technology.

Stent placement in a coronary artery initiates an inflammatory response that includes mobilization of progenitor cells and in-growth of smooth muscle cells. Patients with DM have accelerated neointimal hyperplasia and high rates of subsequent restenosis after stent placement.¹⁸ Neointimal hyperplasia results from convergent mechanisms including increased TGF-ß signaling and a direct influence of hyperglycemia on smooth muscle cell migration.¹⁹ Restenosis is more common after placement of longer stents or in arteries with smaller diameters, and patients with DM tend to have longer, more complex coronary artery lesions, and smaller reference vessels.²⁰ All of these factors converge on higher rates of restenosis when compared to patients without DM.

Stent thrombosis, defined as a sudden thrombotic occlusion of the stent resulting in presentation with an acute coronary syndrome or sudden death, can occur any time after stent placement. It is most common within the first 30 days after stent implantation (early stent thrombosis), but can occur from 31 days to 1 year (late stent thrombosis) or more than 1 year after stent implantation (very late stent thrombosis).²¹ Diabetic patients have increased platelet activation and adhesion, as well as a more prothrombotic neointima after stent placement.²² In recent clinical trials, patients with DM have between 25% and 80% increased risk of stent thrombosis when compared to non-DM patients.^23,24 For these reasons, patients with DM may in particular benefit from more intensive antiplatelet therapy after stent placement, as discussed below. Continued improvements in stent technology may also reduce the long-term risks of stent thrombosis after PCI (see below).

Drug-Eluting Stent Technologies in Patients With Diabetes

DES comprise a bare metal stent platform, a linking or coating mechanism for controlled drug release, and an antiproliferative agent that elutes from the stent over the course of weeks to months.¹⁶ Each technological feature of DES has undergone iterative development over the past decade, including use of thinner and more flexible stent struts, development of more biocompatible and biodegradable polymers (outside of the United States), and newer antiproliferative agents.^25,26 Although strut thickness and polymer type are important determinants of outcomes, DES are categorized primarily by the antiproliferative agent, as summarized in Table 1. With each iteration, DES have in general demonstrated improved safety profiles, with associated lower rates of restenosis, stent thrombosis, and target vessel failure relative to previous designs.

Table 1.

Current FDA-Approved Drug-Eluting Stents.

Variable	Sirolimus-eluting stents	Paclitaxel-eluting stents		Everolimus-eluting stents			Zotarolimus-eluting stents
Commercial name	Cypher	Taxus Express	Taxus Liberte	Xience	Promus	Promus Element	Endeavor	Resolute
Manufacturer	Cordis, Johnson & Johnson	Boston Scientific	Boston Scientific	Abbott Vascular	Boston Scientific	Boston Scientific	Medtronic	Medtronic
Platform material	Stainless steel	Stainless steel	Stainless steel	Cobalt-chrome	Cobalt-chrome	Platinum-chrome	Cobalt-chrome	Cobalt-chrome
Strut thickness (µm)	140	132	97	81	81	81	91	91
Type of polymer	Durable	Durable	Durable	Durable	Durable	Durable	Durable	Durable
Drug concentration (µg/cm2)	140	100	100	100	100	100	160	160
Drug release	80% during first 30 days	10% during first 10 days	10% during first 10 days	80% during first 30 days	80% during first 30 days	80% during first 30 days	80% during first 10 days	80% during first 60 days

Reproduced, with permission, from Stefanini and Holmes.¹⁶

Sirolimus-Eluting and Paclitaxel-Eluting Stents

Sirolimus-eluting stents (SES) and paclitaxel-eluting stents (PES) were the first DES approved by the FDA. Although SES are still used in other countries, this stent type is no longer available in the United States. The PES stent has undergone incremental improvement, with subsequent development of the Taxus® Liberte stent, which has thinner stent struts compared to the first-generation Taxus® DES.

Studies of both SES and PES have demonstrated efficacy in patients with DM consistent with the effect on the overall population.^27-29 Some studies of SES stents suggested a late hazard of death among diabetic patients during 5-year follow-up, possibly related to an increased incidence of very late stent thrombosis.³⁰ However, other studies did not show any relationship between SES stent placement and increased mortality in patients with DM.³¹ Direct comparisons of PES and SES stents in patients with diabetes suggested that PES stents are associated with higher rates of stent thrombosis in diabetic patients.³²

Everolimus-Eluting Stents

Everolimus-eluting stents (EES) were approved for clinical use after PES and SES. Two companies manufacture EES, with 1 stent type derived from cobalt-chromium and the other from a platinum-chrome alloy; both stent types have identical strut thickness and drug elution kinetics (Table 1). Studies in patients with DM demonstrated that EES are associated with decreased neointima formation and angiographic lumen loss when compared to PES.³³ EES are also associated with decreased rates of stent thrombosis when compared to either PES or SES DES.^33-35 In comparison, analysis of other trials suggested that EES do not have significant benefit in diabetic patients when compared to other second-generation DES (Figure 1).^36,37 These findings need to be confirmed in future studies dedicated to enrollment of diabetic patients, with careful designation of the insulin-requiring status of patients with DM.

Figure 1.

Differential outcomes after percutaneous coronary intervention (PCI) among patients with and without diabetes treated with everolimus-eluting or paclitaxel-eluting stents. (A) Among patients without diabetes, the rates of major adverse cardiovascular events were significantly lower for those treated with everolimus-eluting stents (EES) compared to paclitaxel-eluting stents (PES). However, patients with DM had similar outcomes regardless of the drug-eluting stent type. (B) Among patients without diabetes, EES were also associated with significantly decreased rates of target lesion revascularization (TLR). In comparison, there was no relationship between stent type and TLR among patients with DM. (Reproduced, with permission, from Stone et al.³⁶)

Zotarolimus-Eluting Stents

Zotarolimus-eluting stents (ZES) consist of a cobalt-chrome platform, a durable biocompatible polymer, and zotarolimus. The Endeavor® ZES elutes zotarolimus over a short time frame (80% elution over the first 10 days), while the more recent Resolute® ZES has a different polymer and longer drug elution time (80% elution over the first 60 days).

Initial studies of the Endeavor ZES showed similar clinical outcomes compared to PES, but with increased restenosis.^38,39 Subsequent registries and comparative registries confirmed that the Endeavor ZES was associated with higher rates of restenosis and target vessel revascularization in diabetic patients when compared to other DES types.⁴⁰ More recent studies with the newer Resolute ZES have demonstrated improved clinical outcomes compared to the initial Endeavor ZES, and similar outcomes in both diabetic and nondiabetic patients when compared to EES.⁴¹

As part of the FDA approval process, the Resolute ZES was also studied specifically in diabetic patients. Based on a meta-analysis of DES in patients with DM, a performance goal of 14.5% was set for 12-month target vessel failure. At 12 months, the prespecified cohort of diabetic patients had an event rate of 7.8%.⁴² Based on these results, the Resolute ZES was approved for use in patients with DM, and is the only DES with a specific indication for use in diabetic patients. Importantly, the event rates among patients with DM were driven primarily by patients with insulin-requiring DM. Patients with DM that could be adequately treated with oral agents had event rates that were not significantly different from non-DM patients (Figure 2). These findings emphasize that insulin-requiring patients with DM are the highest risk subset of patients undergoing coronary artery revascularization.

Figure 2.

Association of insulin-treated diabetes with outcomes after percutaneous coronary intervention with Resolute zotarolimus-eluting stents (R-ZES). Patient-level data were pooled from 5130 patients receiving R-ZES in 5 clinical trials (Resolute FIM, Resolute All Comers, Resolute International, Resolute US and Resolute Japan). Patients without diabetes (non-DM) and those with diabetes treated with agents other than insulin (non-ITDM) had similar rates of target lesion failure (TLF; defined as a composite of cardiac death, target vessel myocardial infarction, and ischemia-driven target vessel revascularization). Patients with diabetes who required treatment with insulin (ITDM) had significantly increased rates of adverse events. (Reproduced, with permission, from Silber et al.⁴²)

Future Developments in Drug-Eluting Stent Technologies

A number of technological opportunities exist to further improve the outcomes of DES. These include innovations to the stent scaffold, the drug-eluting polymer, and the antirestenotic drug. Although none of these design iterations are targeted specifically at diabetic patients, clinical improvements in stent outcomes should proportionally benefit diabetic patients as well.

Current generation stent designs have thin stent struts (under 100 µm) that maintain radial force within the vessel and prevent acute recoil during the healing phase after balloon angioplasty. A bioresorbable stent could provide the necessary scaffolding to prevent acute recoil immediately post-angioplasty but minimize the risk of late events.⁴³ A number of materials have been used to engineer bioresorbable stents, including magnesium, poly-L lactic acid (PLLA), and polycarbonate. Initial iterations of these devices were hampered by increased late lumen loss relative to permanent metal scaffolds, but newer devices have increased radial strength, thereby minimizing acute recoil.⁴⁴ An everolimus-eluting bioresorbable stent based on a PLLA scaffold is currently approved for use in Europe and has recently entered clinical trials in the United States.⁴⁵ In diabetic patients, the application of bioresorbable stent technology could have the potential to minimize the risk of late target lesion failure and restore normal physiology within the stented region. The results of ongoing trials will determine whether using a bioresorbable stent improves clinical outcomes.

The polymer that links the stent to the eluted drug determines drug elution pharmacokinetics. After the drug is eluted locally over the course of weeks to months, the continued presence of polymer may act as a nidus for chronic inflammation and possibly predispose to stent thrombosis. Newer generation stents contain more biocompatible polymers, but ideally the polymer would degrade after completion of drug elution. A number of stents with biodegradable polymers (primarily based on PLLA) have been studied in small clinical trials, and some of these stent designs are available outside of the United States.⁴⁶ In general, these studies have found that polymer-free stents may be superior to first-generation DES, but there have not been any studies directly comparing polymer-free DES to second-generation DES.

New developments in nanotechnology may also obviate the need for polymer-based drug elution. Currently available polymer-free technologies elute drug over a faster time period and therefore may not provide the longer duration of drug necessary to reduce restenosis compared to currently available polymer-based DES.⁴⁷ However, future developments in this area could allow healing comparable to bare metal stents with the advantage of inhibiting restenosis.

Drugs that inhibit the target of rapamycin (mTOR) pathway (eg, everolimus, zotarolimus) are the most frequently used agents in newer generation DES. An unmet need remains for development of agents that could specifically interrupt the signaling pathways in diabetes that predispose to restenosis. Coupling such an agent to a DES could result in a stent system designed specifically for use in diabetic patients. Further preclinical research in this area could lead to design of a stent that is optimized for use in diabetic patients.

Medical Treatment in Patients with Diabetes and Recent Percutaneous Coronary Intervention

After PCI, optimal medical therapy is crucial to prevention of secondary disease progression and minimizing the risk of stent failure.⁴⁸ Despite the well-recognized role of secondary prevention, a significant percentage of diabetic patients do not meet guideline-recommended goals of therapy after revascularization.⁴⁹ All patients post-PCI should receive aspirin and statin medications indefinitely. Among diabetic patients, other pertinent medical considerations include antiplatelet therapy, the role of intensive glucose control, and the possible effects of metformin on subsequent outcomes.

Antiplatelet Therapy

Dual antiplatelet therapy (DAPT) after PCI is crucial to minimize the risk of stent thrombosis. While the risk of stent thrombosis is highest in the first 30 days after PCI, a persistent risk of stent thrombosis remains, likely due to impaired re-endothelialization, malapposition, or neoatherosclerosis within the stent.⁵⁰

For many years, clopidogrel was the antiplatelet agent of choice for DAPT. Recently, prasugrel and ticagrelor were approved as newer antiplatelet agents that have more predictable metabolism and increased inhibition of platelet activation relative to clopidogrel. Because patients with DM have activated platelets and increased atherothrombosis relative to nondiabetic patients, these agents may have an important role in improving outcome post-PCI in diabetic patients.

Studies of prasugrel in patients undergoing PCI for acute coronary syndromes demonstrated 2.2% absolute reduction in major adverse cardiovascular events, and a 0.6% absolute increase in major bleeding events.²⁴ For every 1000 patients treated, these results translate into prevention of 23 myocardial infarctions at the expense of 6 additional major bleeding events. Patients with DM had a greater net clinical benefit than the overall study population, with a 4.8% absolute reduction in major adverse cardiovascular events and an increase in bleeding risk similar to the overall population.⁵¹ Rates of stent thrombosis were also significantly reduced among diabetic patients.

Ticagrelor was also studied in a large population of patients with acute coronary syndromes, including patients managed both conservatively and with PCI. Compared to clopidogrel, ticagrelor was associated with an absolute risk reduction of 1.9% for major adverse cardiovascular events, without a significant increase risk in overall bleeding events.²³ In patients with DM, the reduction in major adverse cardiovascular events was similar to the overall population, with a 2.1% absolute risk reduction.⁵² There was also a trend toward a reduction in stent thrombosis among diabetic patients.

These studies suggest that diabetic patients with acute coronary syndromes represent a high-risk cohort of patients who may derive additional clinical benefit from more intensive antiplatelet therapy. These considerations need to be balanced against a possible increase bleeding risk. In addition, prasugrel should be administered at a reduced dose (5 mg) in patients weighing < 60 kg and is relatively contraindicated among patients older than 75 years of age. Ticagrelor must also be administered with low-dose (≤100 mg) aspirin, due to a possible association between higher aspirin doses and increased mortality with ticagrelor.

Intensive Glucose Control

Intensive glucose control lowers the risk of macrovascular cardiovascular disease among Type 1 diabetics, but recent studies of intensive glucose control among patients with long-standing type 2 DM have failed to show a significant benefit of achieving an HbA1C ≤ 7.0%.^53,54 After PCI with DES, the data remain mixed on the benefit of intensive glucose control. Some studies, especially during the era of bare metal stents, found that elevated HbA1C levels were associated with an increased risk of restenosis.⁵⁵ In comparison, more recent studies have failed to find a relationship between HbA1C and restenosis, even when HbA1C was analyzed as a continuous variable.⁵⁶ In all recent studies of DES, however, patients with DM requiring insulin have had higher subsequent adverse cardiovascular events compared to patients with DM not requiring insulin. Whether this greater risk reflects more advanced DM among the insulin-requiring cohort or a direct effect of insulin therapy remains uncertain.

Metformin

Metformin is a biguanide that has been used for many years as an oral antidiabetic agent. Some studies have suggested a possible role for metformin in prevention of cardiovascular disease among patients with DM.^57,58 In the setting of patients with established CAD who undergo PCI, metformin may impair endothelial cell recovery and re-endothelialization due to a convergent signaling mechanism with mTOR, the cellular target of everolimus, sirolimus, and zotarolimus. These findings were recently observed in a rabbit model of stent placement and re-endothelialization, suggesting that metformin could be associated with an increased risk of stent thrombosis after DES placement.⁵⁹

It remains unclear whether these signaling interactions of metformin with -limus eluting drugs translate into clinical outcomes. We therefore performed a preliminary analysis of diabetic patients receiving metformin who underwent placement of R-ZES from the RESOLUTE All-Comers trial. At 30 days and 1-year follow-up, there was no significant difference in cardiac death or target vessel myocardial infarction (MI) among patients treated with metformin (Table 2). We are currently analyzing data from a larger pooled data set with a longer follow-up period from other DES trials to determine whether metformin is associated with adverse outcomes with respect to late and very late stent thrombosis among diabetic patients post-PCI.

Table 2.

Demographics and Outcomes of Diabetic Patients Treated With Metformin.

	DM patients with metformin (n = 132)	DM patients without metformin (n = 136)	Adjusted P valuea
Patient characteristics
Age (years)
Mean ± SD (n)	65.4 ± 10.0 (132)	68.0 ± 9.3 (136)	.030
Male	68.9% (91/132)	73.5% (100/136)	.406
Prior myocardial infarction (MI)	28.1% (36/128)	38.9% (51/131)	.066
History of smoking	50.0% (66/132)	51.5% (70/136)	.810
Current smoker	20.5% (27/132)	16.2% (22/136)	.365
Hyperlipidemia	77.3% (102/132)	80.9% (110/136)	.467
Diabetes mellitus	100.0% (132/132)	100.0% (136/136)	–
Insulin dependent diabetes Mellitus	22.7% (30/132)	48.5% (66/136)	<.001
Hypertension	90.2% (119/132)	86.0% (117/136)	.298
Serum creatinine (umol/l)
Mean ± SD (N)	82.7 ± 29.0 (129)	100.0 ± 49.5 (135)	<.001
Range (min, max)	(40.0, 221.0)	(44.2, 420.0)
Safety measures (to 30 days)
Death	0.8% (1/132)	0.0% (0/134)	.362
Cardiac death	0.8% (1/132)	0.0% (0/134)	.362
Cardiac death or target vessel MI (TVMI)	4.5% (6/132)	3.0% (4/134)	.784
Stent thrombosis (ARC definite/probable)	1.5% (2/132)	1.5% (2/134)	.866
Safety measures (to 360 days)
Death	4.5% (6/132)	3.0% (4/134)	.436
Cardiac death	4.5% (6/132)	2.2% (3/134)	.157
Noncardiac death	0.0% (0/132)	0.7% (1/134)	.326
Cardiac death or target vessel MI (TVMI)	9.1% (12/132)	6.7% (9/134)	.456
Stent thrombosis (ARC definite/probable)	3.0% (4/132)	2.2% (3/134)	.823

Data were derived from the Resolute All Comers Trial.

^a P values were adjusted after propensity weighting for differences in baseline demographic characteristics.

Conclusions and Future Directions

Significant advances have been made over the past decade at improving the outcomes of patients with DM and CAD. Large, randomized studies specifically focused on diabetic patients have emphasized the diffuse nature of atherosclerosis among these patients and the increased rates of subsequent cardiovascular events even after revascularization and optimal medical therapy. The development of DES has significantly improved the outcomes of diabetic patients undergoing PCI. Patients with diabetes treated with agents other than insulin now have event rates post-PCI similar to that of patients without DM. In comparison, diabetic patients requiring insulin remain a high-risk cohort despite advances in stent technologies and medical therapy. Careful consideration should be given to more intensive antiplatelet therapy post-PCI among diabetic patients to minimize the risk of atherothrombosis. New technologies on the horizon, including bioresorbable stents and polymer-free drug elution, may provide additional tools to improve the outcomes of patients with diabetes and CAD.