What do we know about the use of DES versus bare-metal stents (BMS) in diabetics with CAD?
To compare DES and BMS in diabetics, you first have to review the issues that diabetics have that non-diabetics only partially share. With either stent, the Achilles' heel of coronary artery stenting is restenosis. Patients with longer lesions and smaller vessel diameters both diabetic and non-diabetic have a higher risk of restenosis. However, because of metabolic considerations, the incidence of restenosis is even higher in diabetics. Data from the control arm of the TAXUS IV Trial indicate the risk of restenosis in patients with similar lesion length and vessel diameter was increased in diabetics compared with non-diabetics.

We have to remember that DES have not been proven safe and effective in diabetics. There is data that we can analyze, but a definitive conclusion has not been made. That said, if one looks at DES in any population of patients, there are two questions to ask. First, do DES appear to offer any advantage over BMS in reducing restenosis or its clinical surrogate, target lesion revascularization (TLR)? Second, do DES and BMS have similar safety data in the diabetic population? There is data that points to 'yes' for both questions, but again, nothing has been proven. TAXUS meta-analysis studies suggest that diabetics do not have an increased risk of stent thrombosis with DES versus BMS, although the diabetic population may have an increased risk of stent thrombosis because of the lesions associated with diabetes.

What specific data is relevant in evaluating DES utilization in diabetics?
While it is unproven, there is evidence available to examine. For example, the TAXUS IV Trial was one of the first studies to show a significant reduction in restenosis with DES as well as a similar reduction in diabetic patients. At nine months, the TLR rate for all patients was about 3 percent. For diabetics, it was a bit higher 5.2 percent in medically treated diabetics and approximately 5.9 percent in insulin-dependant diabetics. This was shown to be a durable effect in follow-up data.

The problem with looking at diabetic data from pivotal trials such as the TAXUS IV Trial and the SIRIUS Trial is that these studies were not powered to look at subgroups such as diabetics. The percentage of diabetics in both studies was about 25 percent, a relatively small number of patients. Another way to evaluate the effects of DES in diabetics is by looking at the meta-analysis of the TAXUS II, IV, V and VI Trials, which included about 3,400 patients, over 800 of whom were diabetic. The data from this meta-analysis show 7.9 percent TLR rates in diabetics on oral agents and 5.8 percent on insulin-requiring diabetics versus 5.4 percent for non-diabetics. Based on these numbers and trials, the diabetic patient receiving a DES experienced an approximately 60 percent reduction in TLR versus BMS.

Are there any other trials that provide insight into the use of DES in diabetics?
I am most interested in TLR. My patients do not ask about their expected late loss or binary restenosis rate; they want to know their chance of having to return in nine months to one year for a second intervention. The ISAR-DIABETES Trial has garnered a lot of attention in the interventional cardiology community, probably because it is the only study comparing the efficacy of the Cypher and TAXUS Stents specifically in a diabetic population. The ISAR-DIABETES Trial shows a significant reduction in late-lumen loss and angiographic restenosis at six months with the Cypher Stent compared to the TAXUS Stent. Essentially all available studies demonstrate that late lumen loss is higher with the TAXUS Stent when compared to the Cypher Stent. When you look at the secondary endpoint of TLR, there is a numerical advantage for the Cypher Stent, with a TLR of 6.4 percent versus 12 percent for the TAXUS Stent. However, the difference in TLR between the two stents was statistically insignificant (p=0.13). The ISAR-DIABETES Trial is a single-center study that was underpowered to show differences between the Cypher and TAXUS Stents, and shows an insignificant difference in TLR rates.

The REALITY Trial, a large, multi-center study which also compares the efficacy of the Cypher and TAXUS Stents, included 490 diabetics. TLR rates for this trial were 5.0 percent for the Cypher Stent and 5.4 percent for the TAXUS Stent, a statistically insignificant difference. If I had to hang my hat on one set of existing data, I would look at the TLR rates in the diabetic meta-analysis of the TAXUS series and the REALITY Trial.

Do you think either one of the DES currently on the market has an advantage in diabetics?
Several different cellular pathways mediate the process of restenosis. It is theorized that there may be at least one pathway that is unique to the diabetic, and this may be why diabetics have a higher incidence of restenosis than non-diabetics. There is interesting research at the basic science level suggesting Paclitaxel may be a more effective drug than sirolimus in blocking this unique pathway but, as a clinician, I have to rely on clinical data. At this time, we do not have studies powered to show an advantage of one stent over the other in terms of efficacy in diabetics. I don't think I can give a leg up clinically to either stent in diabetics. As I said earlier, my patients are not as interested in the data points, but whether or not they will be back within a year for a re-intervention. The fact is, there is still much to learn about DES in diabetics, but at the end of the day, it is the patient that benefits from all of the discussion and evaluation of this method of treating CAD.

Cypher is a trademark of Cordis Corporation.
Sponsored by Boston Scientific.