Background Vascular graft infection(VGI)is one of the dreadful complications invascular surgery. Once it happens, its outcomes (death, limb loss, recurrentinfection) were very poor whether it is treated with conservative or surgicalmethods. In recent years, vascular graft infection as a thorny problem hascaught attention because of the widespread use of vascular graft, and thedevelopment of infection-resistant graft has become the subject of study.Studies have shown that antibiotic-bonded graft is effective in resisting VGI inthat it can obtain a high concentration of antibiotics not only on the prosthesisbut also in the periprosthetic tissues for a period of time. However, there are still problems that should be solved: the bondings donot last long enough and the antibiotics are not effective against mostorganisms involved, especially their drug-resistant strains. The ideal bonding should have long duration that is enough to preventboth early and late VGI, and the ideal antibiotic should be active against mostorganisms that are responsible for VGI. In order to develop ideal infection-resistant graft, we explored thebonding techniques and properties of antibiotics, and tried to prepare thespecific antibiotic-bonded graft, and then designed both in vitro and in vivoantimicrobial tests [using methicillin-resistant S.epidermidis (MRSE) andmethicillin-resistant S.aureus (MRSA) as test organisms] to evaluate itseffectiveness in preventing VGI .Methods:1. Preparation of norvancomycin-bonded Dacron prostheses. 1) An appropriate HPLC method for the drug loading and in vitro release of norvancomycin of DV-DP was established. 2) Norvancomycin-polylactic acid microspheres (NV-PLA-MS) were prepared by emulsion and solvent evaporation method. 3) Norvancomycin-bonded Dacron prostheses were prepared by bonding the NV-PLA-MS onto the surface of Dacron prostheses with solvent evaporation method using ethylene-vinyl acetate (EVA) copolymer as an anchor. 4) The morphology of norvancomycin-bonded Dacron prostheses was evaluated by macroscopic and microscopic examinations. 5) The drug loading of norvancomycin-bonded Dacron prostheses was determined by HPLC. 6) In vitro release of norvancomycin-bonded Dacron prostheses was determined by HPLC. 7) Dichloromethane residue in norvancomycin-bonded Dacron prostheses was determined by GC.2. Measurement of in vitro antibacterial activity of NV-DP 1) With the use of MRSE and MRSA as test organisms, culture of the mixture of a piece of 6-mm- diameter round Dacron prostheses(after placing in 10ml phosphate-buffered saline in a rotary shaking water bath at 370C for 0, 7, 14, 28 days )and 10ml broth containing 0.5ml bacterial suspension was performed. 2) The colony-forming units (CFU) of staphylococcus were then plate-counted after culturing. 3) The colony-forming units from norvancomycin-bonded Dacron prostheses cultures were compared with those from unprotected Dacron prostheses cultures.3. NV-DP in the prevention of methicillin-resistant staphylococcus endovascular graft infection. 1) Dog model was established in the infrarenal aorta of dog by placement of endovascular stent graft (constructed from unprotected Dacron prostheses or NV-DP and a self-expanding stent) followed by topical contamination with MRSE or MRSA at a concentration of 2 x 107 CUF/ml. 2) The samples were scheduled to harvest at 14 days after implantation, and Macroscopic, bacteriologic and histological evaluation of the stent grafts and perigraft vessel samples was performed. The end point of the investigation was graft infection. 3) Comparison of graft infection was performed between norvancomycin-bonded grafts and unprotected grafts.Results1. The mean diameter and drug loading of NV-PLA-MS were 30.7μm and 0.2533±0.0