Design-Dependent Variations in Coronary Stent Stenosis Measured as Precisely by Angiography as by Histology

BACKGROUND: Coronary stent restenosis is a growing clinical concern which, because restenosis may vary with stent design, requires a validated, accurate, and sensitive method of evaluation as new stents are developed. Histologic analysis of arterial cross sections, a highly accurate tool in animal models, has limited applicability in humans. Quantitative coronary angiography, while commonly used in the clinical evaluation of coronary interventions, has a controversial role as an adequate measure of restenosis, and few studies have validated quantitative angiography in diseased arteries. We tested the hypothesis that in-stent stenosis could be assessed as accurately by pre-mortem angiography as by post-mortem histology, allowing angiographic discrimination of variable late luminal loss provoked by stents of different designs. METHODS AND RESULTS: Stent stenosis in porcine coronary arteries was assessed by quantitative coronary angiography and histology at 3, 28 and 56 days. Four stainless steel stent designs were studied: a slotted tube configuration with or without a polymer wrap and a corrugated ring configuration with or without a polymer coating. Although acute luminal gain (mean stent:artery ratio 1.07 +/- 0.01) and stent recoil (mean stent diameter at follow-up 2.65 +/- 0.02 mm) were similar for all designs and time points, significant differences in late luminal loss were observed and were detected as accurately by angiography as by histology. At 28 days, the polymer wrapped slotted tube design resulted in a nearly two-fold greater late loss than its bare metal counterpart (1.40 +/- 0.09 mm vs. 0.80 +/- 0.12 mm, p <.001 by angiography; 1.33 +/- 0.10 mm vs. 0.67 +/- 0.06 mm, p <.0001 by histology), while there was no significant difference in 28 day late loss between the polymer coated and bare metal corrugated ring designs (p = NS by angiography or histology). Time point differences were also observed both angiographically and histologically, with marked progression of lumen loss between 3 and 28 days and slower but persistent progression between 28 and 56 days. Overall comparison of individual lumen diameter measurements for all stented arteries independent of design or duration of follow-up demonstrated a precise correlation between angiography and histology (y = 0.96x +0.25, p <.0001, r2 = 0.82). CONCLUSIONS: Coronary stents of varied designs provoke markedly different degrees of late luminal loss, and these differences can be measured as accurately by quantitative angiography as by histologic analysis of arterial cross sections. This may be due to optimization of angiographic measurements by the concentric nature of intimal thickening in stented arteries, and to structural rigidity imparted by the stent, preserving arterial lumen size for histologic analysis. Quantitative angiography, therefore, may represent an adequate endpoint in clinical trials comparing stent designs.