Background: Heparin inhibits vascular smooth muscle cell proliferation in tissue culture and limits neointimal hyperplasia after experimental arterial injury but has been ineffective in reducing clinical restenosis. We examined how this discrepancy might reflect suboptimal drug-tissue interactions and/or differences in the vascular response to injury.
Methods and results: Intravenous infusion was compared with local administration of heparin to injured rabbit iliac arteries either from drug-impregnated polymeric controlled release matrices in the perivascular space or from drug-releasing endovascular stents. Occlusive thrombosis, seen in 42% of control stent-bearing arteries, and partial thrombosis were virtually eliminated by heparin delivery from any route. Intimal area 14 days after balloon withdrawal denudation alone was reduced to an equal extent by continuous systemic heparin or by perivascular heparin for the first 3 days. In contrast, endovascular stents produced more exuberant neointimal hyperplasia, the inhibition of which required continuous rather than only early heparin administration. Neither perivascular delivery limited to the first 3 days nor stent-based delivery reduced neointimal hyperplasia as effectively.
Conclusions: The antiproliferative and antithrombotic effects of heparin differ markedly, depending on the type of arterial injury and the mode of drug administration. Different forms of injury may require different therapies, and complications of arterial intervention such as excessive neointimal hyperplasia and thrombosis may demand alternate therapeutic regimens. Duration, dose, and site of delivery rather than frank resistance to therapy may explain why experimentally effective antiproliferative and antithrombotic agents fail clinically.