Purpose: This study was undertaken (1) to determine whether correction of infrainguinal bypass stenoses detected with color duplex scanning (CDS) improved graft survival and (2) to define the natural history of grafts that did not undergo revision.
Methods: Over a 39-month period 462 color-flow duplex scans were obtained on 170 limbs with autogenous vein grafts. Grafts were scanned within 3 months of operation, at 6 and 12 months, and then yearly. Doubling of the velocity at any point in the graft-arterial system compared with the velocity immediately above or below (velocity ratio > or = 2.0) was the criterion adopted for identification of a hemodynamically significant (> or = 50%) diameter reduction.
Results: One hundred ten stenoses were detected in 62 (36%) of the limbs, of which 9 (8%) were in native vessels, 30 (27%) were at the anastomoses, and 71 (65%) were in the graft itself. Seventy-seven percent of the stenoses were detected in the first year. Twenty-four (39%) of the grafts with positive scans were revised. During follow-up, occlusions occurred in 10 (9%) of the 108 grafts with negative scans (NEG), in 2 (8%) of the 24 revised grafts with positive scans (PR), and in 10 (26%) of the 38 non-revised grafts with positive scans (PNR). Cumulative patency rates of NEG grafts were 90% at 1 year and 83% at 2 through 4 years. Similar patency rates were found in the PR vein grafts: 96% at 1 year and 88% at 2 through 4 years. In contrast, patency rates in PNR grafts with 50% or greater stenoses were only 66% at 1 year and 57% at 2 through 4 years. Log-rank tests showed a significant difference between the cumulative patency rates of NEG and PNR grafts (p < 0.002) and between PR and PNR grafts (p = 0.02). Flow velocities less than 45 cm/sec and ankle/brachial indexes did not discriminate well between grafts with or without 50% or greater stenoses or identify those grafts that subsequently occluded.
Conclusions: The results of this study suggest that CDS detects graft-threatening lesions, that a velocity ratio of 2.0 or greater is the most highly predictive parameter, and that revision of grafts with stenoses identified with CDS prolongs patency.