BACKGROUND: We have shown that extracellular matrix (ECM) rather than cell proliferation contributes to in-stent restenosis. Transforming growth factor-beta (TGF-beta), a positive regulator of ECM deposition by vascular cells, may be implicated in in-stent restenosis. We assessed if the blockade of TGF-beta by catheter-based local delivery of an adenovirus expressing a soluble form of TGF-beta type II receptor (AdTbeta-ExR) can inhibit stent-induced neointima. METHODS: AdTbeta-ExR was applied onto a coronary arterial segment of a pig using an Infiltrator(TM), and either of adenovirus expressing beta-galactosidase (AdLacZ) or PBS was applied onto other remote segment of the same pig (n=10). Then, stents (n=20) were deployed in the treated arterial segment. RESULTS: Computer-based morphometric analysis 4 weeks after stenting showed no significant difference in neointima area between the AdTbeta-ExR-infected and control groups (AdLacZ and PBS). However cell density of neointima was significantly increased in the AdTbeta-ExR group compared with control group (3121+/-331 vs 2812+/-183 cells/mm2, p<0.05). Notably, the AdTbeta-ExR group had more extensive CD3 positive T cell infiltration. In addition matrix metalloproteinase (MMP)1 expression and accumulation of hyaluronan was greater in the AdTbeta-ExR group. Cell proliferation rate was significantly increased in the media of the AdTbeta-ExR group compared with control group (2.04+/-1.21% vs 1.18+/-1.06%, p<0.05). CONCLUSION: Blockade of TGF-beta by use of catheter-based local in vivo gene delivery did not alter neointima formation significantly in our porcine coronary artery stent model, however it increased inflammation and pathological changes that could promote lesion formation.