Arteriovenous Fistula, Blood Pressure, and Shunt Flow: A Thin Line That Separates Beneficial From Detrimental Effects

Arteriovenous fistula (AVF) is used for initiation of renal replacement therapy in patients with end-stage chronic kidney disease. The creation of an AVF exposes the circulation to significant hemodynamic changes as a consequence of fluid redistribution from the high-pressure, low-capacitance arterial to the low-pressure, high-capacitance venous system leading to reduced peripheral resistance, increased preload, associated with potential right and left ventricular remodeling and myocardial dysfunction. Despite the fact that more than 1.5 million patients in Europe and the United States suffer from end-stage renal disease (ESRD), comprehensive data on invasive and noninvasive hemodynamic assessment, including systemic blood pressure (BP) and pulmonary pressures following AVF creation are scarce.1,2 In this issue of the Journal, Matthew et al. provide valuable insights into the effects of AVF creation on BP in pre-dialysis.3 The study included a relatively large patient population of 399 patients with AVF, which were propensity score matched to 798 patients receiving a catheter for initiation of hemodialysis (HD) and showed a modest but significant reduction of BP in AVF patients. A major strength of the analysis is the utilization of Veterans Affairs Informatics and Computing Infrastructure, thereby ensuring that all BP measurements were recorded before initiation of HD, which can affect the measurements. The present study, however, used in-office nonstandardized BP measurements only, which are known to be less accurate when compared with out-of-office BP readings.4 In general, cautious evaluation of BP and a clear distinction of BP measurements to the timing of HD sessions are essential to allow appropriate interpretation of the recordings in patients with AVF. In fact, the hypothesis of a clinical meaningful BP reduction following AVF creation is not new; clinical observations documented hemodynamic changes post-AVF surgery almost 100 years ago.5 The overall research interest on this topic was, in part, reignited by the ROX Control HTN study, which investigated the consequences of a central AVF creation using a 4-mm stent device (ROX Medical Inc., San Clemente, CA) in patients with severe hypertension and preserved renal function.6–8 In this open-label, multicenter, prospective, randomized, controlled study significant reductions were documented in mean 24-hour ambulatory BP by 13.5/13.5 mm Hg and office BP by 26.9/20.1 mm Hg at 6-month follow-up in patients treated with the coupler with no secondary BP rise during 12-month follow-up.6–8 To accumulate evidence, we performed and published a systematic meta-analysis and sought to illustrate the hemodynamic consequences following AVF creation and ligation, based on all available data.9 However, data research revealed that there is only a limited number of valid studies with moderate quality mostly including small patient populations. Extensive literature research including different databases (PubMed, Web of Science, and the Cochrane Library) identified a total of 14 eligible studies (a total of 1,239 articles were excluded) including 412 patients with ESRD, in which BP was sufficiently measured and risk of potential bias, especially with clear relationship to HD, was acceptable. In these patients, mean BP with no/closed AVF was 140.5/77.6 mm Hg (mean arterial BP was 96.1 mm Hg) and decreased significantly by 8.7/5.9 mm Hg (mean BP decreased by 6.6 mm Hg), when AVF was created. To substantiate these BP reductions, we studied the population where an AVF was ligated and BP tended to increase after AVF ligation, underlining the plausibility of the observed effects. Interestingly, the magnitude of the observed effect documented in the analysis by Matthew et al. was modest with a BP reduction of 1.7/3.9 mm Hg, which differs from the results of our meta-analysis and also from the ROX Control HTN study.6–9 As patient characteristics were homogenous for patients with pre-ESRD the smaller magnitude of the observed effect might be related to the deviation in observational time (12 months) when compared with our meta-analysis.9 One may further speculate that differences in baseline BP, which was 143/76 mm Hg in the present study, significantly higher in the ROX coupler group (175/100 mm Hg), and comparable in our meta-analysis (140/78 mm Hg), account in part for the observed disparancies. It is also important to recognize that the disease states were quite different (pre-ESRD vs. ESRD vs. treatment resistant hypertension) ]with distinct levels of arterial stiffness in these populations, as progression of chronic kidney disease is associated with severe arterial calcification.10,11 It is plausible that patients with severely elevated arterial stiffness benefit most from volume unloading caused by AVF creation.12 However, valid data on arterial stiffness were not available, as these assessments are not routinely performed in clinical practice. Last, the magnitude of the BP reduction in the present study when compared with other published evidence (Figure 1) may be related to differences in shunt flow, which was demonstrated to correlate with the observed systolic BP reduction in AVF.9 Unfortunately, data on shunt flow were not provided in the present study. Correlation of mean systolic blood pressure (mm Hg) and shunt flow (ml/min). Values are expressed as mean. ROX was included and shunt flow calculated with 1000 ml/min; modified from Scholz et al.9 HFS, high-flow shunt; HAIDI, hemodialysis-induced distal ischemia; r2, coefficient of determination. In light of the large number of patients scheduled for AVF creation, reports on complications, i.e., right heart failure caused by the increased venous return/ventricular work, and left ventricular remodeling, are important. Observed side effects associated with AVF include venous stenosis, leg swelling, increased pulmonary pressure, and right heart dysfunction, which are induced by a pathological venous outflow.6–8,13–15 These side effects may prompt closure of AVF in symptomatic patients. It has nicely been shown by invasive hemodynamic assessment (pressure volume loops) before and immediately after interventional AVF creation in severe hypertension that the volume shift from the high-pressure low-capacitance arterial to the low-pressure high-capacitance venous system results acutely in an increased cardiac output and stroke volume.16 Especially long-term effects of this artificially increased volume loading to the right heart are of relevance. Reddy et al.17 recently published a retrospective study of 137 ESRD patients who underwent AVF creation and documented a significant right heart dilatation and deterioration of right heart function, which caused incident heart failure in more than 40% of the included patients. In contrast, it has been shown that creation of an AVF can modestly reduce left ventricular size and mass, and may delay progression of chronic kidney disease in certain patients.17 This is in line with the results of the present study that documented a significantly slower decline of estimated glomerular filtration rate in the AVF group in this pre-ESRD population. The present study adds further evidence concerning the hemodynamic effects of AVF on BP. Further investigations are necessary to confirm whether the BP reduction following AVF creation for initiation of HD are clinically relevant and whether they are preserved in long-term follow-up (>12 months). This may, however, be challenging because most patients will undergo HD, which further affects BP, after a certain period of time. Detailed assessment of shunt flow is necessary to determine an optimal shunt diameter, where BP reduction is superior to an increased risk due to right ventricular stress and volume overloading of the right ventricle.12 Owing to the large number of patients with ESRD requiring HD treatment worldwide, precise evaluation of the hemodynamic changes and rational evaluation of individual advantages and disadvantages following AVF creation are important. Indeed, the issue by Matthew et al. provides additional evidence on BP changes following AVF creation in a relatively large pre-ESRD patient subset and provides a relatively long follow-up period. SSS has received speaker honoraria from Pfizer. MB and FM are supported by Deutsche Forschungsgemeinschaft (SFB TRR219). MB receives honoraria for lectures and scientific advice from Abbott, Astra-Zeneca, Boehringer-Ingelheim, Medtronic, Novartis, Servier, and Vifor. FM is supported by Deutsche Hochdruckliga (DHL) and Deutsche Gesellschaft für Kardiologie (DGK). FM has received scientific support and speaker honoraria from Medtronic and ReCor Medical.