Frequency of critical stenosis in primary arteriovenous fistulae before hemodialysis access: should duplex ultrasound surveillance be the standard of care?

Jennifer Grogan, Maria Castilla, Laurie Lozanski, Andrea Griffin, Frank Loth, Hisham Bassiouny, Jennifer Grogan, Maria Castilla, Laurie Lozanski, Andrea Griffin, Frank Loth, Hisham Bassiouny

Abstract

Objective: Increasing use of primary arteriovenous fistulae (pAVFs) is a desired goal in hemodialysis patients (National Kidney Foundation /Dialysis Outcome Quality Initiative guidelines). However, in many instances, pAVFs fail to adequately mature due to ill-defined mechanisms. We therefore investigated pAVFs with color duplex ultrasound (CDU) surveillance 4 to 12 weeks postoperatively to identify hemodynamically significant abnormalities that may contribute to pAVF failure.

Methods: From March 2001 to October 2003, 54 upper extremity pAVFs were subjected to CDU assessment before access. A peak systolic velocity ratio (SVR) of >/=2:1 was used to detect >/=50% stenosis involving arterial inflow and venous outflow, whereas an SVR of >/=3:1 was used to detect >/=50% anastomotic stenosis. CDU findings were compared with preoperative vein mapping and postoperative fistulography when available.

Results: Of 54 pAVFs, there were 23 brachiocephalic, 14 radiocephalic, and 17 basilic vein transpositions. By CDU surveillance, 11 (20%) were occluded and 14 (26%) were negative. Twenty-nine (54%) pAVFs had 38 hemodynamically significant CDU abnormalities. These included 16 (42%) venous outflow, 13 (34%) anastomotic, and 2 (5%) inflow stenoses. In seven (18%), branch steal with reduced flow was found. In 35 of 54 (65%) pAVFs, preoperative vein mapping was available and demonstrated adequate vein size (>/=3 mm) and outflow in 86% of cases. Twenty-one fistulograms (38%) were available for verifying the CDU abnormalities. In each fistulogram, the arterial inflow, anastomosis, and venous outflow were compared with the CDU findings (63 segments). The sensitivity, specificity, and accuracy of CDU in detecting pAVF stenoses >/=50% were 93%, 94%, was 97%, respectively.

Conclusions: Before initiation of hemodialysis, an unexpectedly high prevalence of critical stenoses was found in patent pAVFs using CDU surveillance. These de novo stenoses appear to develop rapidly after arterialization of the upper extremity superficial veins and can be reliably detected by CDU surveillance. Turbulent flow conditions in pAVFs may play a role in inducing progressive vein wall and valve leaflet intimal thickening, although stenoses may be due to venous abnormalities that predate AVF placement. Routine CDU surveillance of pAVFs should be considered to identify and correct flow-limiting stenoses that may compromise pAVF long-term patency and use.

Source: PubMed

3
購読する