Neointimal hyperplasia after sirolimus-eluting and paclitaxel-eluting stent implantation in diabetic patients: The Randomized Diabetes and Drug-Eluting Stent (DiabeDES) Intravascular Ultrasound Trial

Aims: Patients with diabetes have increased risk of in-stent restenosis after coronary stent implantation owing to neointimal hyperplasia (NIH). The aim of the study was to evaluate the extent and distribution of NIH with intravascular ultrasound (IVUS) after coronary artery stenting with sirolimus-eluting (Cypher) or paclitaxel-eluting (Taxus) stents in diabetic patients. Methods and results: One hundred and thirty diabetic patients were randomized to Cypher or Taxus stent implantation. IVUS was performed at 8 month follow-up. NIH volume was significantly reduced in the Cypher group when compared with the Taxus group: median (inter-quartile range) 0.0 (0.0-0.0) vs. 8.0 mm³ (0.1-33.0), P < 0.001. Per cent NIH volume was also significantly lower in Cypher stents compared with Taxus stents: median (inter-quartile range) 0.0 (0.0-0.0) vs. 7.5% (0.1-27.0), P < 0.001. NIH was covering 5.4% of the stent length in the Cypher stents compared with 46.1% in the Taxus stents (P < 0.001). The incidence of diffuse NIH was significantly higher for Taxus than for Cypher stents (42.9 vs. 3.5%, P < 0.001). Taxus stents had more often NIH at the proximal stent edge compared with Cypher stents (45.1 vs. 7%, P < 0.001) and no Cypher stents had NIH at the distal stent edge compared with 35.5% of the Taxus stents (P < 0.001). Conclusion: In diabetic patients, the Cypher stent, compared with the Taxus stent, inhibited NIH more effectively and had a more focal NIH pattern including less involvement of the stent edges.