An autopsy series of patients with drug-eluting stents gives new insight into late stent thrombosis, including possible identification of risk factors, according to a presentation at the annual meeting of the European Society of Cardiology.

The stents currently approved in the USA (CYPHER and TAXUS) have been associated with late stent thrombosis (thrombosis more than 30days following implantation). Then current autopsy study was designed to analyze morphologic changes occurring after placement of drug-eluting stents to determine the causes of late stent thrombosis.

Researchers autopsied 83 patients (117 lesions) with coronary artery disease who had drug-eluting stents in place for more than 30 days prior to death. A total of 33 lesions showed either luminal thrombus (25 lesions) or organized thrombus (8 lesions).

Of the 117 total lesions, 6 (in five patients) showed a hypersensitivity reaction (5 Cypher, 1 Taxus).The six stents had been in place from 112 to 940 days; three patients died suddenly and two presented with acute myocardial infarction. From these data it appears that response is limited to the area of the stent, and there is extensive eosinophilic and T-cell infiltration. There may or may not be a granulomatous reaction.

Other morphologic changes that predisposed to stent thrombosis were malapposition (8 patients), stenting of bifurcation lesions (7 patients), acute myocardial infarction (8 patients), and overlapping stents (4 patients). All showed delayed healing, which was further exaggerated either from turbulent flow at malapposition or bifurcation sites or poor healing at sites of plaque rupture or overlapping stents.

Excessive length (more than 30 mm) was a correlate of late thrombosis as well as presence of uncovered stent struts. Uniformly, all cases with thrombi had the presence of fibrin, poor stent coverage by neointima, and less endothelialization.

All 78 lesions that were patent (and drug-eluting stent was not the cause of death) and had been in place for more than 30 days showed less neointima compared with bare metal stents, suggesting that drug-eluting stents are effective in reducing neointimal thickness. A parameter uniformly observed in bare metal stents is that neointimal formation around the circumference of the stent tends to be uniform in distribution. In drug-eluting stents, there is heterogeneity of healing with areas showing excessive fibrin and others with smooth muscle cells within matrix and uneven luminal endothelialization.

Both Cypher and Taxus stents reduced neointimal formation from delayed healing. However, there were inherent differences in the response to each stent, with fibrin deposition more frequent in Taxus stents and inflammation, especially eosinophilic infiltrate and giant cell reaction, greater in Cypher stents.