Abstract Background and Aims Vascular access (VA) guidelines recommend radio-cephalic (RC) over upper arm autogenous arteriovenous fistulas (AVF) as first line VA for hemodialysis in end stage renal disease (ESRD) patients. RCAVFs generally have inferior maturation and patency rates predicated on a lower feeding arterial blood flow (BF) and outflow vein calibre (VC). However studies on postoperative BF and VC as predictors of AVF outcomes, so far are confounded by their focus on early outcomes only, heterogeneity of AVFs studied, variable timing of assessment and use of non-standardised outcome definitions. Our aim was therefore to assess the accuracy and influence of immediate post-operative BF and VC on both early and longterm outcomes in a homogenous cohort of primary RCAVFs using standardised definitions and outcome measures as mandated by VA guidelines. Method This was a prospective study conducted in multi-ethnic Asian ESRD patients who had their primary RCAVFs created between October 2013 and October 2014 under regional anesthesia at Khoo Teck Puat hospital Singapore. All AVFs were assessed immediately after surgery for brachial artery BF and outflow VC using doppler ultrasound. A 10MHz linear probe and GE Logic e R7 machine were used exclusively by a single operator. Receiver operating characteristic (ROC) curves were generated to determine the optimal BF and VC cut-off for AVF maturation. Maturation was defined as BF>600mL/min, VC>6mm and vein depth <6mm at 6 weeks post-op. An area under the curve (AUC)> 0.7 was considered clinically significant. Kaplan–Meier analysis was used to evaluate the AVF primary and secondary patency based on best BF and VC cut-offs. Cox regression statistics was used to determine AVF hazard factors. Results Fifty-seven primary RCAVFs were created and included in the study. The baseline characteristics are shown in Table 1. Sonography- based non-assisted maturation at 6 weeks was 56%. ROC identified 410 mL/min and 42mm as the best BF and VC cut-off respectively to most accurately predict 6-week maturation. The sensitivity, specificity, positive predictive value and negative predictive value were 75%, 61%, 44% and 86% for BF at 410 mL/min and 69%, 61%, 41% and 83% for VC at 42mm respectively. Survival analysis (Fig. 1 and 2) showed that AVFs with VC≥42 mm compared to <42mm had significantly greater 6 months, 1-year, 2-year and 4-year primary and secondary patency rates. There was no significant difference in patency rates between AVFs with BF≥410 and <410mL/min. Cox proportional regression hazard analysis showed that diabetes (HR 2.26, CI 1.02-4.99, p= 0.04) and maturation (HR 0.47, 95% CJ 0.24-0.89, p=0.02) as significant contributed to the variability of primary patency. Only VC (HR 0.28, 95% CI 0.13-0.063, p=0.002) impacted significantly towards secondary patency. Conclusion An immediate post-op BF≥410mL/min and VC≥42mm can predict early RCAVF outcome in the form of nonassisted maturation, but only VC accurately impact on longterm AVF survival. VA surveillance efforts should therefore target RCAVFs with post-op VC <42mm for timely intervention and maintenance of longterm patency.
Read full abstract