JACC in a Flash | Diabetes and Lesion Complexity

For patients undergoing percutaneous coronary revascularization (PCI), diabetes mellitus (DM) can increase the risk of angiographic restenosis and ischemia-driven target lesion revascularization (TLR) and target vessel revascularization (TVR). While this is particularly true with bare-metal stents (BMS), drug-eluting stents (DES) have significantly reduced complication rates compared to BMS in patients with and without diabetes. Does that mean DES have effectively eliminated DM as a risk factor for restenosis? According to Elvin Kedhi, MD, PhD, and colleagues, the answer may depend on lesion complexity.

To determine the impact of DM on patient outcomes as a function of baseline lesion complexity, Kedhi et al. analyzed data from a large patientlevel pooled-database of 18,471 DEStreated patients from 18 prospective randomized trials. Medically-treated DM was present in 3,467 (18.8%) of 18,441 patients whose DM status was known. A match was found for 3,167 DM patients, and thus a total of 6,334 patients comprised the analysis population (mean age = 63 years). Lesion severity, defined by the ACC/AHA classification, and lesion length were well matched between the two groups, although diabetic vessels were slightly smaller.

As shown in the TABLE, there were no significant differences in any endpoint according to diabetic status at 30 days. One-year outcomes, however, revealed a trend towards worse outcomes in patients with DM. Multivariable analysis revealed that DM was an independent predictor of TLR, TVR, cardiac death or MI, and major adverse cardiovascular events (MACE; p < 0.0001 for all).

TABLE. Clinical Outcomes at 30 Days and 1 Year, According to Diabetic Status

30-Day Outcomes


Diabetes Mellitus (n = 3,167)

No Diabetes Mellitus (n = 3,167)

HR (95% CI; p Value)


0.7% (22)

0.4% (12)

1.84 (0.91-3.71; 0.08)

Myocardial infarction

2.8% (82)

2.4% (75)

1.09 (0.80-1.49; 0.58)

Stent thrombosis

0.8% (26)

0.5% (15)

1.74 (0.92-3.28; 0.08)

Target vessel revascularization

1.2% (38)

1.2% (30)

1.27 (0.79-2.05; 0.33)

Cardiac death or MI

3.4% (101)

2.6% (82)

1.23 (0.92-1.65; 0.16)

Major adverse cardiac events

4.0% (119)

3.4% (99)

1.20 (0.92-1.57; 0.18)

1-Year Outcomes


2.6% (79)

1.4% (42)

1.91 (1.32-2.78; 0.0005)

Myocardial infarction

4.1% (126)

3.2% (100)

1.27 (0.98-1.65; 0.07)

Stent thrombosis

1.4% (42)

0.7% (21)

2.02 (1.20-3.42; 0.007)

Target vessel revascularization

9.4% (284)

6.2% (191)

1.54 (1.28-1.85; < 0.0001)

Cardiac death or MI

5.3% (164)

3.8% (120)

1.38 (1.09-1.75; 0.007)

Major adverse cardiac events

13.9% (429)

9.4% (294)

1.50 (1.29-1.74; < 0.0001)

When data were stratified by lesion complexity, Kedhi et al. found that TLR and TVR rates were increased in type B2/C compared to simpler A/B1 lesions in the DM cohort, but not in the non-DM cohort. A significant interaction was found between DM and ACC/AHA lesion type for the 1-year rates of TLR (pinteraction = 0.01) and TVR (pinteraction = 0.02). Rates of cardiac death or MI and MACE were also consistently higher in patients with compared to those without DM; lesion severity did not play a role in this interaction. Safety and efficacy interactions between DM status and ACC/AHA lesion complexity were consistent when the lesion types were analyzed separately.

"The present analysis, representing the largest study to date examining DES outcomes in patients with and without DM, demonstrates that DM remains an independent predictor of adverse safety and efficacy outcomes in the DES era," Dr. Kedhi and colleagues concluded, but this association is particularly strong in DM patients with complex lesions, when repeat revascularization after DES was significantly more common than in non-DM patients.

Ultimately, these data suggest that PCI might have favorable results compared to coronary artery bypass graft surgery (CABG) if the extent of disease is not great. Both cardiac death and MI rates were, however, increased in DM patients, regardless of lesion complexity, "emphasizing the need for earlier detection and more effective systemic therapies in this high-risk condition."

Unfortunately, as Stephen G. Ellis, MD, pointed out in an accompanying editorial, patients with diabetes often have more advanced coronary artery disease (CAD) than non-diabetics when they present for revascularization. So, how should a clinician choose between CABG and PCI for a patient with CAD and DM who is technically suitable for either approach?

According to the conclusions drawn by Kedhi et al., it might well be that diabetics presenting with "simple" anatomy might fare just as well with PCI as bypass surgery, Dr. Ellis noted: "It seems that the availability of current DES has mitigated the difference between diabetics and non-diabetics with only a few 'simple' lesions for local lesion recurrence, but not for overall death or myocardial infarction." But this victory needs to be tempered, he concluded, "A step forward, yes, but a small one as stenting does not alter the aggressive panvascular aspect of diabetes."

Ellis SG. J Am Coll Cardiol. 2014 February 28. [Epub ahead of print]
Kedhi E, Genereux P, Palmerini T, et al. J Am Coll Cardiol. 2014 February 28. [Epub ahead of print]

Clinical Topics: Cardiac Surgery, Invasive Cardiovascular Angiography and Intervention, Aortic Surgery, Interventions and Coronary Artery Disease

Keywords: Coronary Artery Disease, Myocardial Infarction, Drug-Eluting Stents, Risk Factors, Coronary Artery Bypass, Percutaneous Coronary Intervention

< Back to Listings