#2851 Effects of dialysis on various microcirculatory beds in end-stage kidney disease (ESKD): a systematic review and meta-analysis

Abstract

Abstract Background and Aims Microcirculatory dysfunction is common in ESKD-patients and may play a pivotal role in their cardiovascular burden. Besides traditional and chronic-kidney-disease (CKD)-specific risk factors, dialysis itself might contribute to the dysfunction of the microcirculation (MC). Although an intradialytic fall in systemic blood pressure (BP) may directly influence the microcirculatory perfusion, vascular beds in different organs may not react uniformly. This systematic review and meta-analysis evaluated the acute impact of hemodialysis (HD) and hemodiafiltration (HDF) on the MC in various vascular beds. Methods Pubmed and Embase databases were systematically explored until June 2023 by two independent researchers. All observational and interventional studies in ESKD patients that reported dialysis-induced alterations of the MC, measured at the cutaneous, sublingual, retinal, muscle or coronary beds were eligible. Dialysis-induced changes were extracted and pooled, using a fixed-effects meta-analysis when the studies were reasonably homogeneous in terms of outcomes and methodology. Results 17 Studies with 493 participants were included. Virtually all patients (99.8%) were treated with HD. No eligible studies on coronary or muscle MC were identified. Cutaneous MC was investigated in 7 studies, showing a mean reduction in transcutaneous oxygen pressure of −7.9 mmHg (95% CI −10.5 to –5.3, n = 4 studies) and a mean reduction in blood flow of −0.9 arbitrary units (95% CI −1.2 to −0.6, n = 3 studies). Cutaneous vasoreactivity was diminished in 2 out of 3 studies (reactive maximum microvascular hyperemia: −36% (p < 0.05); time to reach baseline skin temperature after standardized cooling: 65.6 seconds (p-value not reported); vs. no significant change in post-ischemic maximal flow (2.5, p = not significant (N.S.)). Sublingual MC was assessed in 3 studies, showing a mean decrease in microvascular flow index of −0.6 in small microvessels (vessel diameter (VD) <20-25 µm, 95% CI −1.0 to −0.2, n = 3 studies), but no significant change in proportion of perfused vessels (−0.5%, 95% CI −1.2 to 0.2, n = 2 studies). One study showed a significant reduction in sublingual perfused vessel density after treatment (VD <20 µm: −1.6 ± 3.1, p < 0.05; VD >20 µm: −0.6 ± 1.1, p < 0.05), whereas total vessel density was unaffected in another study (VD <20 µm: 0.5, p = N.S.). Retinal MC was investigated in 7 studies, showing that the perfused vessel density (PVD) remained unchanged in both the superficial and deep capillary plexus (PVD SCP −0.48, 95% CI (−0.99; 0.03); PVD DCP −0.48, 95% CI (−1.18; 0.23), n = 4 studies). One study reported a decreased PVD at the outer retina (−1.3 ± 5.7, p < 0.05), while conflicting results were observed in 2 studies (0.1, p = N.S.; vs. −2.9 ± 4.3, p < 0.001), reporting the PVD at the choriocapillaris level. One study found no change in retinal blood flow (0.4, p = N.S.). With regard to vessel diameter changes, an increase in retinal venular diameter was observed in 3 studies (retinal venular diameter: 11.2 µm, p = 0.002; central retinal vein equivalent: 12.7 µm, p < 0.001; retinal vessel diameter: 4.4, p < 0.01), and one study described impaired venular dilatation in response to flicker light (−4.0, p = 0.04). Conclusions Hemodialysis induces microvascular dysfunction in different organs, but its impact varies between vascular beds. Microvascular perfusion was most significantly impaired in the cutaneous and sublingual regions, whereas the retinal microcirculatory perfusion remained unaltered. Literature on the effect of hemodiafiltration on the MC is lacking.