Membrane Bioincompatibility and Ultrafiltration Effects on Pulse Wave Analysis during Haemodialysis

Membrane bioincompatibility was demonstrated by successive white blood cell counts and C3a generation. Pulse wave analysis was obtained by applanation tonometry (SphygmoCor) in a sequential way: basal, after 30 minutes with nul ultrafiltration, and after a complete dialysis with ultrafiltration. At 15 minutes of haemodialysis, significant decrease in leukocyte count occurred: 6801 ± 1186 versus 4412 ± 1333 (P < 0.001), while C3a levels sharply increased from 427 ± 269 to 3501 ± 1638 ng/mL (P < 0.000). No changes were observed in augmentation index without ultrafiltration: 26.1 ± 11.1 versus 26.6 ± 12.4. Only aortic systolic blood pressure was lower at 15 minutes: 120.1 ± 17.7 versus 110.4 ± 25.8 mmHg (P = 0.009), in agreement with a reduction in brachial systolic blood pressure: 135.1 ± 18.1 versus 122.7 ± 27.4 mmHg (P = 0.01), without changes in aortic or brachial diastolic blood pressure. Important changes in pulse wave analysis were observed after a complete haemodialysis session: augmentation index 29.9 ± 10.1 versus 18.6 ± 15.0, aortic systolic blood pressure 139.8 ± 25.5 versus 119.4 ± 28.5 mmHg (P < 0.00), without changes in aortic diastolic blood pressure. In summary, haemodialysis with cellulose diacetate acutely induced a transient state of immunoactivation due to bioincompatibility, this phenomenon was nondetectable by pulse wave analysis. Complete haemodialysis session led to important changes in pulse wave analysis.


Introduction
Cardiovascular (CV) disease is the main cause of morbidity and mortality in patients with end-stage renal disease (ESRD). e increased risk is partly due to a higher prevalence of traditional CV risk factors. Nevertheless, these patients also present other nontraditional CV risk factors related with the setting of uremic background [1] that result in functional and structural alterations of the arterial wall, leading to an increase in arterial stiffness. As in the general population, arterial stiffening has been described as an independent predictor of both CV and overall mortality in haemodialysis (HD) patients [2][3][4][5][6][7][8].
Another component that has been proposed to play a role in CV risk is the HD session per se. e dialysis procedure may induce acute functional alterations of the arterial wall through several mechanisms, the most remarkable being the intermittent immunoactivation state induced by dialysis [9][10][11], and the acute intravascular volume drop produced during the HD session [12][13][14][15][16][17]. It has been proposed that these acute functional alterations could be detected by noninvasive pulse wave analysis (PWA) [12].
As PWA measurements are being increasingly introduced in the clinical setting, the main objective of our study was to separately analyze the acute effects on PWA of bioincompatibility and ultra�ltration (��) during the dialysis session.   (1) they had been on HD for at least 3 months, (2) they were in a stable clinical situation, and (3) they did not suffer cardiovascular instability on dialysis. All selected patients agreed to participate in the study, which was approved by the Hospital Ethics Committee.

Pulse Wave Analysis (PWA).
Radial arterial waveform in the non�stula arm was measured by applanation tonometry with the SphygmoCor device, using the PWA soware package. Studied variables included augmentation index corrected by heart rate (AI@75), subendocardial viability ratio (SEVR), ejection duration (ED), and aortic systolic and diastolic blood pressure. Measurements were taken by a single observer in triplicate and averaged. UF effects results are displayed in Table 2. Important changes in all the PWA parameters evaluated were observed at the end of the haemodialysis session: AI 29.9 ± 10.1 versus 18.6 ± 15.0 (Figure 1), ED 37.6 ± 3.6 versus 32.8 ± 4.6, SEVR 124.6 ± 19.9 versus 171.7 ± 37.1, central aortic systolic BP 139.8 ± 25.5 versus 119.4 ± 28.5 mmHg (all 0.00), without changes in aortic diastolic BP 75.8 ± 13.9 versus 75.1 ± 17.9 mmHg (p : ns). However, there was no good correlation between the amount ultra�ltered and the changes in PWA parameters ( 0.65), suggesting that arterial stiffness improvement was not caused by �uid removal alone.

Discussion
Arterial stiffness measurement in ESRD patients has gained importance in the last years, especially for its relationship with increased le ventricular hypertrophy and decreased coronary perfusion during diastole [12]. Nevertheless, the acute effects of the HD session on arterial stiffness have been sparsely studied. e results of previous publications are controversial, and it is not clear whether HD per se can induce acute functional changes in the arterial wall [18,19]. is study was designed to separately evaluate the two seemingly and opposite main determinants of arterial stiffness changes during the HD session. In Study I, we demonstrate the existence of immunoactivation assessed by considerable changes in leukocyte count and C3 levels. However, such bioincompatibility was not detectable by PWA, since we did not observe any changes in any arterial stiffness parameters. Previous studies have shown changes in vascular function using other functional studies as pulse wave velocity (PWV), �ow-mediated dilation, or endotheliumindependent vasodilation [9,10]. To our knowledge, no previous studies have analyzed the effect of the membrane bioincompatibility on AI@75.
When the dialysis procedure was performed with the normal pattern of UF, and PWA was measured at the end of the dialysis session (Study II), a signi�cant effect on arterial stiffness parameters was shown. We observed a decrease in both aortic systolic BP and AI@75 at the end of the HD session. Contrary to expected, we found no correlation between changes in arterial stiffness parameters and the magnitude of �uid removal. Previous publications have analysed the effect of the HD session on arterial stiffness yielding different results. Some of them defend a lack of improvement of arterial stiffness arguing that vascular changes in ESRD are structural rather than functional [20], that oxidative stress counteracts the effect of UF, or that the bene�cial effect of acute volume reduction may be obscured by the activation of the rennin-angiotensin system [21,22]. Other studies agreed that PWV remains unchanged as a more structural parameter [13,14], while AI@75 decreases with the HD session (as a more functional one). is improvement has mainly been attributed to volume correction [12][13][14][15], although some authors have already questioned this association [16][17][18]. Given the lack of association with the UF, we analysed possible relationships with pre-HD brachial or aortic BP, pre-HD AI@75, BP decrease during HD or dialysis vintage, but we did not �nd correlation with any of these variables.
ese discrepancies in the literature can be �usti�ed by two main reasons. On the one hand, the large number of variables that can in�uence during a HD session, not only immunoactivation and intravascular volume drop, but also BP changes, oxidative stress, activation of the renninangiotensin system, changes in calcium or magnesium levels during the session, or the differences between patients in intravascular re�lling due to individual nutritional status, among others [21][22][23][24][25][26]. It is, therefore, very difficult to analyze each of them separately. On the other hand, there is a lack of uniformity in the way arterial stiffness is measured in the different studies.
We can conclude that HD procedure with the diacetate cellulose membrane induces a clear immunoactivation effect easily demonstrable by leukocyte cell count and C3a generation that is not detectable by PWA. e improvement of arterial stiffness observed aer the HD session assessed by PWA was not related with ultra�ltration nor with changes in BP.
Further studies with control of the different variables should be carried out in order to determine how each of these factors affects arterial stiffness during the hemodialysis session.

�on�ic� o� �n����s�s
None of the authors of the paper have a direct �nancial relation with SphygmoCor Device that might lead to a con�ict of interests.