High-flow arteriovenous fistula and heart failure: could the indexation of blood flow rate and echocardiography have a role in the identification of patients at higher risk?

BACKGROUND: Although only high-flow arteriovenous fistulas (AVFs) are postulated to cause high-output cardiac failure (HOCF), there are currently no universally accepted criteria defining a high-flow fistula.
METHODS: To verify if vascular access blood flow (Qa) ≥ 2000 ml/min provides an accurate definition of high-flow fistula, we selected 29 consecutive patients with Qa ≥ 2000 ml/min at color-duplex ultrasound examination and assessed them for the presence of cardiac failure symptoms; transthoracic echocardiography was also performed.
RESULTS: Nineteen patients (65%) had heart failure symptoms and were classified with HOCF. At receiver operating characteristic (ROC) curve analysis, Qa ml/min values did not identify patients with heart failure symptoms but when AVF blood flow was indexed for height2.7, Qa ≥ 603 ml/min/m2.7 detected the occurrence of HOCF with good accuracy (sensitivity 100%, specificity 60%, efficiency 86%, positive predictive value 83%, negative predictive value 100%, area under curve 0.75). At echocardiographic evaluation, patients with Qa ≥ 603 ml/min/m2.7 had a more severe increase of left ventricular mass (63 ± 18 vs. 47 ± 7 g/m2.7, p < 0.003), left ventricular diastolic volume (140 ± 42 vs. 109 ± 14 ml, p < 0.007), left atrial volume (53 ± 23 vs. 39 ± 5 ml/m2, p < 0.015), a higher incidence of diastolic dysfunction (70 vs. 17%, p < 0.019) and higher CO reduction after AVF manual compression (2151 ± 875 vs. 1292 ± 527 ml/min, p < 0.009) than patients with Qa < 603 ml/min/m2.7.
CONCLUSIONS: Indexation of AVF blood flow should be considered in defining high-flow fistula because the effect of Qa may differ in individuals of different sizes. A Qa value ≥ 603 ml/min/m2.7 and its association with some echocardiographic alterations could identify patients at higher risk for HOCF.