Lipid and Immunophenotypic Profiles in Hemodialysis Patients with Citrate vs. Acetate Dialysates

1. Introduction

Chronic kidney disease (CKD) is a substantial cardiovascular (CV) risk factor, being considered equivalent to having suffered a previous CV event [1,2,3,4]. Patients with dialysis-dependent CKD (DD-CKD) have a 5-year mortality rate of approximately 50%, according to some authors [5,6,7], mainly due to CV causes.

Hypercholesterolemia constitutes another important CV risk factor and is also a therapeutic objective to control in the general population [8,9,10]. In this sense, low-density lipoprotein (LDL) levels are targeted to a certain threshold according to each patient's calculated risk, given the evidence on lipid control as a preventive measure both for primary and secondary CV events in those with high risk, including CKD patients [11,12,13]. The association with mortality and cardiovascular disease between other relevant lipid biomarkers and DD-CKD has been studied in triglycerides (TG) [14,15], high-density (HDL) [16,17,18] and very low-density (VLDL) lipoprotein cholesterol [19,20], and lipoprotein (a) [Lp(a)] [21,22,23], as well as the calculation of the TG/HDL ratio [24].

Unlike other high-cardiovascular-risk populations, the efficacy obtained from lipid-lowering therapy remains uncertain in DD-CKD patients. Two clinical trials and a systematic review failed to prove significant benefits from statins in these patients [25,26,27]; nonetheless, there is data on the subgroup of dialysis patients from a randomized controlled trial and a real-life retrospective study that show a reduction in major cardiovascular events [28,29]. The discrepancy between findings is probably given to the exclusion in clinical trials of patients with very high LDL levels, their population's heterogeneity, accompanying cardiovascular morbidities, and the higher mortality risk explained in part by chronic inflammation [30,31,32]. Interestingly, some studies in patients with DD-CKD have reported that high cholesterol is not associated with decreased mortality but with increased survival [33,34]. These findings suggest that LDL blood levels may indicate malnutrition, inflammation, or a sarcopenic state [33].

There is evidence that lymphocyte subpopulations are linked to CV risk. Studies in non-CKD patients indicate that patients with a CD4/CD8 ratio higher than 1.5 [35] and an increased proportion of natural killer (NK) lymphocytes [36] are at a greater risk of CV problems. Also, the complement system, particularly C3 and the C3/C4 ratio, has been associated with CV disease and metabolic disorders [37,38,39]. It is known that citrate-based dialysates reduce inflammatory biomarkers compared to acetate-based ones, and some authors state that this could be due to reduced oxidative stress and interfere with the immunological inflammation process. Acetate-based dialysates have also been associated with accumulating uremic toxins, membrane biocompatibility, and inflammation [40,41,42].

The objective of this study is to determine how the use of citrate dialysate, compared to acetate, alters the basic lipid profile and immunophenotype of patients with DD-CKD.

2. Materials and Methods

2.1. Study Design and Population

This is a unicentric, cross-over, prospective study. Patients over 18 years old undergoing post-dilution online hemodiafiltration, who have been on dialysis for at least three months, receiving treatment three times a week for at least four hours each session, and maintaining a stable clinical condition during this period, were eligible for inclusion in the study. Each subject underwent 24 dialysis sessions, 12 with each dialysate acidifier, and acted as their own controls. Blood samples were retrieved predialysis on the last session with each acidifier.

All parameters of the dialysis session (calcium, sodium, and bicarbonate prescriptions; blood and dialysate flows; and dialysis duration), along with medical treatments, were kept constant throughout the study, except for the dialysate acidifier. The details of the dialysate characteristics can be found in Table 1. This study utilized Fresenius 6008 CAREsystem™ dialysis monitors and FX CorDiax™ 60 dialyzers (Fresenius Medical Care, Bad Homburg v.d.H., Germany).

2.2. Variables

2.2.1. Lipid Parameters

We registered the lipidic profile by assessing the plasma levels of total cholesterol, LDL, TG, HDL and VLDL lipoprotein cholesterol, and Lp(a), as well as the calculation of the TG/HDL ratio, comparing the use of citrate (CD) vs. acetate (AD) as dialysate acidifier, after twelve sessions with each one.

2.2.2. Immunological Parameters

The studied leucocyte populations were neutrophils, lymphocytes, neutrophil-to-lymphocyte ratio (NLR), natural killers (NK) cells, CD3 positive, CD4 positive, CD8 positive, and CD19 positive count. Complement levels (C3 and C4 levels) and the C3/C4 ratio.

2.2.3. Other Parameters

Blood levels for uric acid, glucose, folic acid, vitamin B12, magnesium, prealbumin, creatin kinase, total proteins, albumin, transferrin, interleukin-6 (IL-6), high-sensitive C-reactive protein (hs-CRP), D-dimer, and erythrocyte sedimentation rate (ESR).

2.3. Statistical Analysis

Quantitative variables are reported with mean and standard deviation when normally distributed or median and 25th and 75th percentiles if skewed. Normal distribution was assessed with the Shapiro-Wilk test, and the comparisons were made with the Student's paired T-test or Wilcoxon's signed-rank test, accordingly. A two-sided p-value ≤ 0.05 was considered statistically significant.

3. Results

Two-hundred and fifty-two sessions were performed with CD and AD in total. Twenty-one participants, consisting of 10 (47.6%) males and 11 (52.4%) females, with an average age of 62.3 ± 13.8 years (ranging from 33.1 to 82.3 years), were included in the study. Dialysis access included arteriovenous fistula for 12 patients and tunneled catheter for 9 patients. Various underlying renal conditions were observed, including diabetic kidney disease (6 patients), autosomal dominant polycystic kidney disease (2 patients), HIV nephropathy (2 patients), hypertensive nephropathy (2 patients), chronic glomerulonephritis (1 patient), chronic pyelonephritis (1 patient), renal cell carcinoma (1 patient), other cystic diseases (1 patient), and undetermined causes (5 patients). Noteworthy comorbidities included cirrhosis with A Child-Pugh score in 1 patient, liver disease with normal function in 5 patients (1 with polycystic hepatorenal disease, 2 with HCV, 1 with HBV, and 1 with HCV-HBV coinfection), and 1 patient who had undergone a liver transplant with normal graft function. Cardiac conditions were present in 3 patients with valve diseases, 2 with chronic ischemic cardiopathy, and 1 with hypertensive cardiomyopathy.

3.1. Lipidic Profile Results

There were significantly lower LDL cholesterol levels with AD compared to CD (AD: 67.05 ± 30.53 v. CD: 75.71 ± 31.01, p= 0.042), higher TG (AD: 110.38 ± 35.79 v. CD: 97.86 ± 29.69, p= 0.046), and higher HDL cholesterol (AD: 51.19 ± 11.25 v. CD: 47.14 ± 11.99, p= 0.013). However, there were no significant differences between predialysis total cholesterol (AD: 140.33 ± 30.56 v. CD: 142.33 ± 33.4, p= 0.624), VLDL (AD: 22.14 ± 7.36 v. CD: 19.67 ± 5.97, p= 0.057), TG/HDL ratio (AD: 2.35 ± 1.16 and CD: 2.29 ± 1.07, p= 0.876), and Lp(a) values (AD: 26.21 ± 30.45 and CD: 27.04 ± 34.21, p= 0.617). See Figure 1 for further details.

3.2. Complement Levels

We found a statistically significant difference in predialysis C3 levels between both dialysates. They were higher when patients were dialyzed with CD than with AD. However, there were no significant differences in C4 levels nor in the C3/C4 ratio (AD: 4.43 ± 0.98 v. CD: 4.5 ± 0.92, p= 0.482). More details in Table 2.

3.3. Leucocyte Count and Subpopulations

There was a higher number of total leucocytes (AD: 7.1 ± 3.9 v. CD: 6.3 ± 2.2, p= 0.03) and lymphocytes (AD: 1.37 ± 0.46 v. CD: 1.21 ± 0.47, p= 0.037) with AD than with CD. We also found a higher percentage of CD3+ (AD: 72.9 ± 12.69 v. CD: 68.48 ± 11.86, p= 0.0049) and CD8+ T cells (AD: 33.43 ± 14.56 v. CD: 30.67 ± 12.85, p= 0.05), and higher CD56+ CD16+ NK cells (AD: 16.95 ± 7.85 vs CD: 19.24 ± 8.84, p= 0.035). There were no statistically significant differences in mean CD4/CD8 ratios nor the proportion of CD4/CD8 >1.5 with each dialysate (AD: 42.1% v. CD: 57.9%, p= 0.352). There were also no differences observed in predialysis CD19 + B-cells, and CD4+ T-cells, total leucocyte count, neutrophils, NLR, monocytes, nor eosinophils between dialysates. See Table 2 for further information.

3.4. Inflammatory and Nutritional Parameters

There were no statistically significant differences in predialysis IL-6 and hs-CRP between dialysates. However, the ESR was higher when patients used CD than AD (AD: 41.19 ± 22.92 v. CD: 53.62 ± 35.24, p= 0.02). Regarding nutritional parameters, we did not find any significant differences between dialysates in predialysis albumin, prealbumin, total proteins, iron, transferrin, magnesium, iron, folic acid, or vitamin B12 levels. More detailed information available in Table 3.

4. Discussion

After twelve dialysis sessions with CD, compared to AD, there was a statistically significant decline in TG and HDL and an increase in LDL. Regarding immunology, C3 complement levels were higher, while CD3+ CD8+ and CD16+56+ lymphocytes were lower. Finally, total lymphocytes were lower with AD than with CD. We found no difference in predialysis nutritional nor inflammatory parameters except for ESR, which was higher when subjects used CD than AD.

Few studies have measured the effect of the dialysates' weak acidifier in the lipidic profile, and when done -always as secondary variables only an increase in LDL been described with the use of CD [43]. Previous studies from the 1980s did not find clinically significant changes in the lipid profile when bicarbonate solutions with reduced concentrations of acetate became commonplace over only acetate solutions [44].

Real-life findings seem to associate lipid values with the opposite of what is expected to be beneficial in the general population. For instance, high TG levels (i.e., > 193 mg/dL) correlate with a lower mortality risk in HD patients [15], while high HDL levels (i.e., > 60mg/dL) appear to correlate with an increased mortality risk in HD patients [18]. In our cohort, the median TG was 101 mg/dL and the highest value registered was 192 mg/dL, not reaching the apparently beneficial value previously published [15], though, it is worth noting that TG levels were significantly higher with AD than with CD. Also, unlike us, a recent study by de Sequera, et al., showed no differences in TG between AD and CD [45]. Regarding HDL, in our cohort we found lower levels with CD than with AD, with 38% of patients having values over 60 mg/dL when using AD while only 14% crossed this threshold while on CD. With regards to LDL, data suggest a U-shaped association with all-cause and cardiovascular mortality risk in peritoneal dialysis patients [46]; however, as previously stated, no clear association has been established with mortality in HD patients.

Further studies must elucidate if the short-term changes induced in the lipidic profile of DD-CKD patients by the dialysate's weak acid translate into clinical implications.

Regarding immunological parameters, there is recent evidence that the use of CD for three months did not affect leucocyte nor total B or T lymphocyte count in comparison to AD in patients on HD [47]. Notably, the same authors measured NLR, which has associated with worse survival in HD patients [48], and, like us, found no differences between dialysates.

Regarding lymphocytic subpopulations, previous studies have shown that there is a U-shaped relationship between the CD4/CD8 T cell- ratio and atherosclerosis progression, being associated with lower ratios in HIV patients (mean ratio of around 0.5) [49,50,51,52] and more than 1.5 in Chinese elderly population (mean ratio of around 1.33) [35]. It is reported that early atherosclerotic plaques have ratios below 1, while late atherosclerotic plaques, particularly in late fibroatheroma, are around 1.5 [53]. We found no differences in the proportion of low and high CD4/CD8 ratios between dialysates. However, our population’s mean CD4/CD8 ratio was higher than that from the Chinese study despite their subjects being on average older than ours. This higher ratio could potentially translate into a higher CV risk, though more research is needed in dialysis patients.

We found higher total leucocytes and lymphocytes with AD than with CD, predominantly by a higher count of CD3+ and CD8+ lymphocytes, whereas NK cells were higher with CD than with AD. The interpretation of lymphocyte subpopulations needs to be clarified. There is evidence that both CD4 and CD8 decline post-hemodialysis, while their predialysis values remain unchanged compared to healthy populations [54]. This is believed to be in response to poor biocompatibility with the dialyzer membrane or even with the dialysis solution, as postulated by data that has associated acetate with an increased number of activated CD3+CD4+CD69+ T cells than with citrate [47].

LDL receptors play a role in CD8+ T cell activation. Studies in ldlr −/− mice and individuals with familial hypercholesterolemia who carry a mutation in LDL receptor show decreased levels of CD8+ cytokine production and cell proliferation [55]. Whether citrate alters LDL receptor function, increasing LDL levels and thus decreasing the number of CD8+ T remains unconfirmed but could partly explain the observed results compared to acetate.

Our cohort found that patients had higher circulating NK cells when exposed to CD than to AD. NK cells constitute another subset of lymphocytes, which may also be necessary in vascular disease [56]. These cells have been isolated from atherosclerotic plaques, particularly expanding the necrotic cores [36]. Regarding circulating levels, a higher percentage of NK cells has been reported in patients with severe atherosclerosis awaiting revascularization [57], elderly patients with coronary disease [58], and has been associated with an increased number of CV and neurological complications after an endarterectomy [59].

Additionally, we found higher TG levels and decreased C3 levels when patients used AD than with CD. These findings may be related, given that data from experimental studies have shown that C3 stimulates glucose uptake and inhibits hormone-sensitive lipase in several cell types [39]. C3 promotes lipophagy, which stimulates VLDL secretion in hepatocytes and balances TG levels in the liver [60]. C3-deficient mice present with glucose intolerance delayed TG clearance and decreased TG storage [39,61,62]. Therefore, C3 has been associated with metabolic disorders and is now recognized as a cardiometabolic risk factor [37]. We also measured the C3/C4 ratio given recent data associating increased serum C3/C4 ratio as a novel marker for recurrent cardiovascular events in acute coronary syndrome [38]. However, we found no differences in this parameter in our cohort.

Multiple sources state lower inflammatory parameters with CD than with AD [40,41,42,63], but there are some discrepancies in the current literature [47,64]. We found no differences in IL-6 or hs-CRP, though we did find a statistically significant difference in ESR in favor of AD. It is worth noting that most of the previous evidence in this regard comes from patients on HD rather than HDF.

When combining our findings, it appears that there may be an increased CV risk when patients use CD. However, previous studies comparing mortality between CD and AD in HD patients either show no difference or suggest a potential benefit for specific subpopulations [65]. This could be because HD patients behave differently than other subjects, or it could be that CD provides a counterbalance that mitigates the risk. It has been attributed that CD potentially reduces vascular calcification, which would be a hypothesis to confirm in further studies.

This study has several limitations that should be considered when interpreting the results. Firstly, the single-center design and the small sample size may limit the generalizability of the findings to other populations of DD-CKD patients. Additionally, the relatively short follow-up period of 24 sessions may not be sufficient to observe the long-term effects of the different dialysates on lipid and immunological profiles. Although the crossover design minimizes interindividual variations, the influence of uncontrolled confounding factors, such as changes in diet, or physical activity, during the study, cannot be entirely excluded. Furthermore, the absence of a non-intervention control group makes it challenging to compare the observed changes with the natural progression of these parameters in the absence of dialysate changes. The measurement of inflammatory and nutritional parameters is also subject to biological and technical variability, which could affect the accuracy of the results. Future multicenter studies with larger sample sizes and longer follow-up periods are needed to confirm these findings and better understand the clinical implications of the observed changes.

5. Conclusions

The use of citrate dialysate in DD-CKD patients leads to significant changes in lipid profiles and immune parameters compared to acetate dialysate. Specifically, citrate dialysate increases LDL and HDL levels while reducing TG and alters various lymphocyte subpopulations. These findings suggest that citrate dialysate may have distinct metabolic and immunological effects, potentially influencing cardiovascular risk and inflammation in DD-CKD patients. Additional research is needed to understand the long-term clinical implications of these short-term changes and to optimize dialysis treatment strategies for this high-risk population. To date, the efficacy of lipid-lowering therapy, atherogenic risk factors, and the mechanisms and impacts of dyslipidemia remain unclear in patients with DD-CKD; thus, new methodological approaches like lipidomics are necessary to address this knowledge gap.