#532 Association between blood flow measured within 30 days from AV access creation and the risk of thrombosis

Abstract

Abstract Background and Aims The use of arteriovenous (AV) access for hemodialysis (HD) is hampered by the risk of dysfunction and thrombosis. Various factors affect the risk of AV access thrombosis, such as age, female sex, hematocrit, vessels’ diameter and anatomy, and patients’ comorbidities (e.g. diabetes mellitus, thrombophilia, chronic inflammatory disease). Previous studies have reported contradictory results on the prognostic value of intraoperative blood flow (Qa) measurements. Finally, only few studies have investigated the association between postoperative Qa measurements and thrombosis risk. Therefore, current guidelines report that there is not enough evidence to suggest routine assessment of AV access Qa for surveillance. This study aimed to analyze the association between Qa, measured at the first post-operative visit (within 30 days from creation), and thrombosis risk, in distal and proximal autologous AV fistulae (dAVF and pAVF), and AV grafts (AVG). Method This retrospective cohort study included all AV accesses created in San Paolo Hospital (Milan, IT) between January 1st 2013 and December 31st 2022, with a post-operative Qa measurement available within 30 days from the date of creation. AV accesses were classified as dAVF, pAVF and AVG. Qa was measured by ultrasound as the product of the cross-sectional area and the time-averaged mean velocity, in the feeding artery. The thrombosis-free survival of dAVF, pAVF and AVG was plotted using the Kaplan Meier method. The association between Qa and the risk of thrombosis was studied by Cox proportional hazard models, adjusted for patient's age and sex. Non-linearity was assessed using natural splines. Results A total of 218 AV accesses (92 dAVF, 76 pAVF, and 50 AVG) were created in 191 patients. At baseline, patients had a median age of 69 years, 70% were male, 42% were already on HD treatment, and overall, they had a high prevalence of hypertension, diabetes mellitus and cardiovascular comorbidities. At first post-operative visit, median Qa was 922 [698, 1278] ml/min, 1260 [917, 1825] ml/min, and 1665 [1195, 2508] ml/min, respectively for dAVF, pAVF, and AVG (p <0.001). During a median follow up time of 1.35 [0.55, 2.62] years, 66 AV access thrombosis occurred. The thrombosis-free survival rates for dAVF, pAVF and AVG were 84%, 97%, and 51% at 1 year, and 63%, 66%, and 19% at 5 years (Fig. 1). For dAVF we found a significant L-shaped association between Qa measured at the first post-operative visit and thrombosis risk, with the lowest risk found at Qa of 1289 ml/min and an exponential risk increase for dAVF with Qa <800 ml/min (Fig. 2). Particularly, for dAVF with Qa < 1289 ml/min, we found a 39% increased risk of thrombosis for every 100 ml/min decrease in Qa (aHR 1.39, 95% CI 1.10-1.75, p 0.006). In pAVF, Qa showed a trend of inverse linear association with the risk of thrombosis, although not reaching statistical significance (aHR 1.06, 95% CI 0.96-1.17, p 0.24, for every 100 ml/min decrease in Qa). In AVG, Qa was not associated with the risk of thrombosis (aHR 1.01, 95% CI 0.97-1.06, p 0.57, for every 100 ml/min decrease in Qa). Conclusion In our cohort we found that Qa, measured at the first post-operative visit, is associated with the risk of thrombosis in dAVF, but not in pAVF and AVG. In particular, in dAVF the association is non-linear with the lowest risk in our cohort found at Qa equal to 1289 ml/min, with an exponentially increasing risk for Qa <800 ml/min. Our findings suggest that post-operative flow assessment may be a useful tool to identify dAVF at higher risk of thrombosis and allow a closer monitoring or possible preventive interventions.