[en] Purpose: Endovascular brachytherapy (EVBT) has been proven to significantly reduce restenosis after percutaneous transluminal angioplasty (PTA). The object of this analysis was to assess the possible correlation between iridium-192 source non-centering and angiographic-determined restenosis. Materials and methods: A total of 113 patients with long-segment lesions of the superficial femoro-popliteal artery (SFA) were randomized to receive either PTA alone or PTA followed by EVBT in the Vienna-2 study. This analysis was performed on a subgroup of 34 out of 57 patients, who received PTA+EVBT. Angiographic restenosis was defined as lumen reduction of more than 50%. Angiograms taken immediately after PTA (34 patients) and at follow-up (25 patients) were analyzed. The distance between the vessel wall and the actual position of the source at the time of EVBT was measured (in mm) and correlated with the follow-up vessel lumen diameter. Measurements were performed at points at a distance of 10 mm from each other. The dose was determined at the luminal surface and at the reference depth of 2 mm into the vessel wall for different distances from the source. Results: Among the 622 measured points, 62 (10.0%) were within restenotic areas; 560 (90.0%) were in arterial segments without proven angiographic restenosis. As far as source centering is concerned, 7.9% of restenotic points were observed when the maximum distance to the arterial wall was <3 mm and 9.6% for 4 mm, respectively. The percentage of restenotic points increased up to 15.9% when the maximum distance to the arterial wall was 5 mm and reached 22.2% when it was >5 mm. Conclusions: The proportion of restenotic points significantly increased with source non-centering. This observation was interpreted as being related to a decrease in dose at the target. When the maximum distance between the source and the vessel surface was >5 mm, the dose at the reference depth (2 mm into the vessel wall) decreased to values lower than 5 Gy