The natural history of stenoses within lower limb arterial bypass grafts using a graft surveillance program.

Graft surveillance aims to identify those grafts that are at risk of failure as intervention in a patent but failing graft results in improved long-term patency and limb salvage rates compared to rescue of an occluded graft. Controversy exists as to which types of graft benefit the most from surveillance and whether patient factors such as diabetes and smoking status have an effect on graft survival. Our aims were (1) to clarify the natural history of midterm graft failure as a consequence of myointimal hyperplasia and (2) to identify which patients and grafts are at a higher risk of failure and at what time points this is most prevalent. Serial vascular laboratory and clinical data of 212 infrainguinal lower limb grafts in 197 patients were analyzed. Follow-up within the surveillance program was by focused examination with color flow duplex ultrasound at 0, 1, 3, 6, 12, and 18 months with respect to surgery. Outcomes were correlated with retrospectively collected data regarding patient demographics, smoking status, concurrent medication, comorbidity, and operative factors such as distal target vessel and conduit. During the program, 21.6% of grafts occluded. Overall, 16% of grafts underwent a salvage procedure, 40.5% of which were carried out at the 6-month time point. There were 56.6% of occlusions preceded by a stenotic lesion. Primary occlusions accounted for 95.9% in the prosthetic group and 66.5% in the femorocrural group. As a group, vein grafts were more likely to develop a progressive stenosis prior to occlusion, with 58.3% in this group predated by a stenotic lesion. Fewer than 75% of stenoses were common and had a variable natural history, with over 40% resolving or failing to progress. Throughout the study period, 56.2% of grafts remained stenosis-free. Stenoses were more common at the proximal anastomosis in the vein graft cohort. There were low rates of significant stenoses within the prosthetic group. These lesions were more likely to occur at the distal anastomosis but were poor predictors of occlusion. Statin use postoperatively was protective against the development of significant stenosis and occlusions, particularly in the above-knee grafts (p = 0.03). Surprisingly, preoperative smoking status was predictive of neither occlusion nor development of significant stenosis. The presence of diabetes was not predictive of poor outcome. Our findings suggest that graft surveillance is a valid method for detecting the presence of significant stenoses in vein grafts at high risk of failure without intervention. Despite the intensive follow-up, the program failed to detect lesions prior to occlusion in a large percentage of prosthetic and femorocrural grafts, so perhaps this group is poorly served by graft surveillance.