EXAMINATION Confirms Poor 10-Year Prognosis in Diabetic Patients Post-STEMI

Diabetic patients who undergo revascularization for STEMI have significantly worse clinical outcomes over 10 years when compared with nondiabetic patients, most often because they require another coronary revascularization, according to long-term follow-up of the EXAMINATION study.

The results of the new analysis, from a study originally designed to compare the safety and efficacy of an everolimus-eluting stent (Xience V; Abbott Vascular) against BMS for the treatment of patients with STEMI, aren’t particularly surprising given the poor prognosis of diabetic patients.

Nonetheless, they do provide some insight into the long-term outcomes of this high-risk population. For the investigators, too, there were some unexpected revelations.

“The thing that most surprised us was that [diabetic patients] had a higher incidence of all-cause mortality, especially in the second part of the follow-up,” said lead investigator Francesco Spione, MD (Hospital Clínic de Barcelona, Spain). Other studies, he added, have demonstrated a higher rate of death among diabetic patients with STEMI in the first 5 years, but this is the first long-term follow-up of such patients, he said.

The subanalysis of the EXAMINATION study was presented as a “Key Abstract” online this week as part of a sneak peek at TCT 2021. Originally published in the Lancet in 2012, the EXAMINATION study was a randomized trial comparing EES against BMS in 1,504 patients with and without diabetes who presented with STEMI. At 1 year, use of Xience did not lower the risk of the primary endpoint (all-cause death, MI, or revascularization) compared with BMS for the overall cohort, but rates of TLR and TVR were lower with the everolimus-eluting stent.

At 5 years, there was a significant reduction in the risk of the primary endpoint among patients treated with Xience, a benefit driven by reductions in mortality and TLR. The 10-year data, which was presented last year at TCT, provided a bit of a twist, in that the event curves for BMS- and EES-treated patients essentially ran parallel between 5 and 10 years. Xience-treated patients had a lower overall risk of death, MI, or revascularization at the 10-year mark, but all of the benefit accrued in the first 5 years.

Revascularization Mostly Within 5 Years

In this latest analysis, investigators focused solely on the clinical outcomes of the 258 diabetic patients included in the trial. Compared with nondiabetic patients, the diabetic patients had a higher risk of the primary composite endpoint in long-term follow-up. “There is a statistically significant difference for the primary endpoint at 10 years,” said Spione, “but if we look at the Kaplan-Meier curves, we can see that the event curves separate in the first year and then run parallel for the entire duration of follow-up.”

There was no overall difference in the risk of all-cause mortality at 10 years between the diabetic and nondiabetic patients, but as Spione noted, the event curves widened between 5 and 10 years in the landmark analysis. From randomization to 5 years, the rate of all-cause mortality was 13.2% and 9.6% in the diabetic and nondiabetic subgroups. Between 5 and 10 years, all-cause mortality in the diabetic patients was 19.6% compared with 10.8% among nondiabetics, a difference that trended toward significance (P = 0.089), said Spione.

Rates of revascularization were also higher in the diabetic patients at 10 years (24.4% vs 16.6%; P = 0.002), but unlike mortality, this risk was driven by events in the first 5 years and by higher rates of revascularization of the nontarget vessel.   

Also of note, said Spione, the worse outcomes in diabetic patients at 10 years were seen regardless of whether they were treated with the everolimus-eluting stent or a bare-metal stent.

Harlan Krumholz, MD (Yale University School of Medicine, New Haven, CT), one of the discussants following the presentation, commended the investigators for the 10-year follow-up, adding that there is not a good understanding of how diabetic patients fare long-term. That said, he added, the new analysis raises some questions.

“One is really about having an idea about what’s going on in those 10 years,” he said. “To what degree is this group, for example, being treated with statins? Are they optimizing their care?” Krumholz said that diabetic patients are a very heterogenous group, running the range from patients who are under tight glycemic control, watch their diets, and take other guideline-recommended therapy to those who aren’t well nearly as well managed. He asked whether any of these factors influenced the clinical outcomes.

Spione said that they don’t have data on glycemic control or background medical therapy in this long-term follow-up analysis, which he conceded was a limitation. Krumholz said not having such data is fine, noting that research always spurs further clinical questions and suggesting this might be addressed with other studies.