Angioscopic observations after percutaneous thrombectomy of thrombosed hemodialysis grafts

PURPOSE: To use angioscopy to evaluate and compare the amount of residual t hrombus and endoluminal wall damage in hemodialysis grafts after percutaneo us thrombectomy procedures. MATERIALS AND METHODS: Thirty-nine thrombectomy and angioscopy procedures w ere performed in 35 patients. Percutaneous thrombectomy methods included ei ght different mechanical thrombectomy devices and the "lyse and wait" techn ique, Videotaped images of 33 angioscopic examinations were independently r eviewed by three radiologists. Two parameters-the amount of residual thromb us and degree of endoluminal wall damage-were scored on a scale of 1 to 5, Data were initially analyzed to validate the grading system and then furthe r studied to compare the different thrombectomy techniques. RESULTS: The Spearman rank order analysis validated the data pertaining to the amount of residual thrombus (r = 0.71, P < .0001), but there was poor c orrelation between reviewers regarding the degree of endoluminal wall damag e. Combined scores from three reviewers revealed that the Cragg brush and P ercutaneous Thrombectomy Device (PTD) left the smallest amounts of residual thrombus, The other methods tested, listed by increasing amount of residua l thrombus, were the Endovac, Hydrolyser, Amplatz Thrombectomy Device, Angi oJet, Oasis, and the lyse and wait technique. There were two complications related to angioscopy procedures. CONCLUSION: Subjective observations reveal that wall-contact thrombectomy d evices leave less residual thrombus than hydrodynamic devices, aspiration d evices, or the lyse and wait technique.