Reduction of Pro-Inflammatory Cytokines through Hemodiafiltration

Abstract

Background. Hemodialysis (HD) prolongs the life of the patients with end stage renal disease (ESRD), but the survival rates are still lower than the general population. More than half of ESRD patients died from cardiovascular disease (CVD). Recent studies have revealed that CVD is a consequence of vascular inflammation, and that there are active inflammatory processes in ESRD patients. Reports have indicated that ESRD patients have fewer CVD events and better survival with hemodiafiltration (HDF), but the reasons for this remain unclear. This study attempts to prove that HDF reduces the CVD-related cytokines. Methods. Seventeen adult HD outpatients were put on HDF in our hospital from September 2004 to June 2006. We collected plasma samples before and six months after initiation of HDF. The target pro-inflammatory cytokines selected were interleukin-6 (IL-6), interleukin-18 (IL-18), tumor necrosis factor-α (TNF-α), and C-reactive protein (CRP). Results. After six months of HDF, most of the biochemical parameters did not changed. Plasma IL-18 and TNF-α are decreased significantly (p < 0.05) but IL-6 and CRP are not. Conclusions. IL-18 and INF-α decreased significantly after six months of HDF. These cytokines are key factors in atherosclerotic plaque formation and rupture, and a reduction of these inflammatory cytokines in HDF may reduce the CVD incidence and prolong life.

Keywords:

• cardiovascular disease
• C-reactive protein
• hemodiafiltration
• interleukin-6
• interleukin-18
• tumor necrosis factor-α

INTRODUCTION

According to 2007's Annual Data Report of the United States Renal Data System, dialysis patients have low survival rates, with a 10-year survival rate at only 12.4%, and more than 50% of mortalities are CVD-related.[1] Recently, CVD has been strongly linked with atherosclerosis, which is a consequence of vascular inflammation.[2],[3]

In ESRD patients, chronic inflammation is a common feature and results in cardiovascular morbidity and mortality.[4],[5] Ozeren et al. reported that serum IL-1β, IL-8, and TNF-α were higher in patients with unstable angina pectoris than controls.[6] In patients with chronic renal failure, plasma IL-6, IL-8, IL-10, CRP, and TNF-α have all been reported higher than the general population.[7–9] Elevated IL-1, IL-6, IL-13 and TNF-α levels are all significantly associated with high mortality rates.[10] Many investigators have found that some pro-inflammatory cytokines can be removed from extracorporeal circulation by convection.[11–13] We designed a prospective study to evaluate the effects of HDF on pro-inflammatory cytokines, which are highly associated with CVD.

METHODS

Patients

After CMUH Institutional Review Board approval and written consent from volunteers, we initiated 17 chronic, stable, adult HD patients to HDF in the HD center of China Medical University Hospital (CMUH), Taiwan, from September 2004 to January 2006. Plasma samples were taken before and six months after the initiation of HDF. All patients had been put on maintenances dialysis for more than three months. They had no evidence of infection, cardiovascular events, hospitalization, surgery, or malignancy in three months. None of the patients were smokers.

HDF

HDF was performed three times per week, in a four-hour session, using a Fresenius 4008S. HDF was applied using single-use, hollow-fiber dialyzers, equipped with polysulfone membranes (F×80 or F×100, Fresenius Medical Care, Germany; APS-18S or APS-21S, Asahi Kasei Medical, China). The membrane surface was 1.8 m² in APS-18S and F ×80, and 2.1 m² in APS-21S and F ×100. The dialysate was bicarbonate-based buffer Hemodialysis Concentrate No. 110 (Taiwan Biotech, Taiwan). Blood flow was 300–350 mL/min. Dialysate flow was 800 mL/min. HDF replacement fluids were generated online and in a post-dilution setting. The exchange volume was 20 liters in each HDF treatment. Vascular access was through forearm arterio-venous fistulas. Heparin was used for anticoagulation; active coagulation time was kept between 200–250 seconds. Dialysate quality met European pharmacopoeia standards (bacterial cultures of the purified water < 100 CFU/mL, and endotoxins < 0.25 EU/mL by sensitive LAL test).

Biochemical Parameters and Cytokines

Fasting blood samples were drawn from the arterial end of the vascular access, immediately before dialysis began. Blood samples were collected in vacutainer tubes containing ethylenediamine tetra-acetic acid. They were then centrifuged at 2000 g for 5 min, with the liquid component retained, and kept frozen at −80°C until assayed. Blood urea nitrogen, creatinine, albumin, total cholesterol, triglycerides, CRP, and other biochemical parameters were measured by standard laboratory techniques with an automatic analyzer in the central laboratory of our hospital. Protein catabolic rate (PCR) in HD patients was calculated by Sargent's method[14] and normalized (nPCR) to actual body weight. Dialysis dose was expressed as Kt/Vurea, in line with Daugirdas.[15]

Plasma IL-6, IL-18, and TNF-α were assayed by a commercially available immunoenzymatic method (Colorimetric Sandwich ELISA; R&D Systems, Minneapolis, Minnesota, USA), according to manufacturer instructions. The detection ranges for IL-6, IL-18, and TNF-α were 3.13–300, 31.2–2000, and 15.6–1000 pg/mL, respectively. The sensitivities were <0.7, <12.5, and <5.5 pg/mL, respectively.

Statistical Analysis

Results were expressed as means ± SEM and were compared using one-way analysis of variance (ANOVA) and the Bonferroni test. Because the cytokine data did not show a normal distribution, differences between groups were examined in a nonparametric way (Mann–Whitney U-tests). Correlation coefficients were calculated through Spearman rank analysis. Differences were considered significant at p < 0.05 for Wilcoxon signed-rank test.

RESULTS

The average age of these seventeen subjects was 49.74 ± 12.92 years. Nine patients were female. Underline kidney diseases were type 2 diabetes (n = 2), chronic glomerulonephritis (n = 9), hypertension (n = 5), and others (n = 1). The average dialysis period was 101.12 ± 267.94 months. Biological and hematological data is listed in Table 1. Most showed no difference after HDF, except in blood urea nitrogen (BUN, 70.47 ± 11.43 vs. 80.71 ± 14.91 mg/dL, p < 0.0001) and aspartate aminotransferase (AST, 15.76 ± 5.82 vs. 18.24 ± 5.83 IU/L, p = 0.0024). Dialysis doses (Kt/Vurea) at initiation and six months after HDF were 2.00 ± 0.31 and 2.02 ± 0.29 (p = 0.3713), respectively. Nutritional status for our patients, presented with nPCR before and six months after HDF, was 1.31 ± 0.22 g/kg/day and 1.32 ± 0.21 g/kg/day, but this was not significant.

Table 1 The biological and hematological data of patients who converted to HDF

n = 17	Before HDF	Six months after HDF	p value
BUN mg/dL	70.47 ± 11.43	80.71 ± 14.91	<0.0001
Creatinine mg/dL	12.53 ± 3.28	13.05 ± 2.37	0.6809
T-Chol mg/dL	173.29 ± 33.14	166.71 ± 33.75	0.2383
TG mg/dL	152.29 ± 192.23	138.21 ± 208.56	0.6593
Albumin g/dL	3.76 ± 0.28	3.74 ± 0.27	0.7754
AST IU/L	15.76 ± 5.82	18.24 ± 5.83	0.0024
ALT IU/L	16.53 ± 5.78	17.59 ± 6.69	0.3296
Uric acid mg/dL	7.03 ± 1.21	7.16 ± 1.54	0.8105
Calcium mg/dL	9.37 ± 0.76	9.49 ± 0.85	0.2942
Phosphate mg/dL	5.44 ± 1.31	5.46 ± 1.18	0.7031
i-PTH pg/mL	216.79 ± 152.06	317.89 ± 183.73	0.0877
Iron μg/dL	54.3518.89	51.88 ± 17.92	0.6776
TIBC μg/dL	239.88 ± 45.54	222.47 ± 56.20	0.1116
Ferritin ng/mL	429.39 ± 245.25	405.84 ± 237.68	0.9265
Hb g/dL	10.71 ± 0.92	10.40 ± .96	0.0750
WBC ×10^3/μl	6.49 ± 1.79	7.05 ± 2.73	0.1861
Platelet ×10^3/μl	190.82 ± 52.04	203.18 ± 62.19	0.2362

Plasma IL-6 did not change significantly after six months (4.19 ± 3.07 vs. 3.28 ± 2.65 pg/mL, p = 0.1816). However, plasma IL-18 was lower than pre-HDF level (344.21 ± 121.80 vs. 267.44 ± 64.90 pg/mL, p = 0.0003), as was plasma TNF-α (9.33 ± 3.45 vs. 8.02 ± 2.15 pg/mL, p = 0.0258). Plasma CRP did not change significantly (0.39 ± 0.23 vs. 0.50 ± 0.50 mg/dL, p = 0.8984; see Table 2).

Table 2 Changes of inflammatory cytokines after six months HDF

n = 17	Before HDF	Six months after HDF	p value
CRP mg/dL	0.39 ± 0.23	0.50 ± 0.50	0.8984
TNF-α pg/mL	9.33 ± 3.45	8.02 ± 2.15	0.0258
IL-6 pg/mL	4.19 ± 3.07	3.28 ± 2.65	0.1816
IL-18 pg/mL	344.21 ± 121.80	267.44 ± 64.90	0.0003

DISCUSSION

There are many active inflammatory processes in ESRD patients; some of which resulting in atherosclerosis and CVD.[16–18] If we can reduce the vascular inflammation, CVD incidence should be lowered. If HDF diminishes vascular inflammations, it would explain why HDF patients have better survival rates. We chose four pro-inflammatory cytokines (CRP, IL-6, IL-18, and TNF-α), all of which are strongly associated with vascular inflammation.

CRP is a risk marker for CVD both in the general population and in ESRD patients.[19] Recently, CRP has been found to be an inflammatory mediator that modulated endothelial cells, macrophages, and vascular smooth muscle cells in atherogenesis.[20] The higher the CRP level is, the greater the CVD risk will be in ESRD patients.[21] However, plasma CRP did not decrease significantly after six months of HDF, which may due to the low CRP baselines.

Interleukin-6 plays a key role in regulating humoral and cellular responses. Many vascular cells, including endothelial, smooth muscle, lymphocytes, and macrophages, can produce IL-6.[22] These cells are highly associated with vascular inflammation. Large quantities of IL-6 have been found in human atherosclerotic plaques,[23] and several studies revealed a direct link between IL-6 levels and CAD.[24] In our patients, there was a trend toward reduced plasma IL-6 levels after HDF, but the differences were not significant.

TNF-α has a major role CAD pathogenesis and has been detected in human atherosclerotic plaques.[25] Increased levels of TNF-α have been observed after acute myocardial infarction.[26] TNF-α regulates the synthesis of fibrinogen and factor VIII, which are risk factors for atherosclerosis.[27] It is heartening that plasma TNF-α was reduced after six months of HDF patients, indicating a reduction of TNF-α-related vascular inflammation. Although previous research has indicated that TNF-α could not be filtrated due to trimer formation in plasma,[28] recent study had shown that it can be removed in high volume CRRT.[29] We believe that TNF-α can also be removed by convection in HDF.

IL-18 was originally identified as interferon-g-inducing factor in Kupffer's cells and macrophages. It also plays a key role in the inflammatory cascade.[30] IL-18 levels have been positively correlated with CRP.[31] Plasma IL-18 levels have been identified as strong independent predictors of death from cardiovascular causes in patients with CVD.[32] Plasma IL-18 is also increased in those with unstable angina and myocardial infarction.[33] High serum IL-18 has been reported in HD patients.[34],[35] IL-18 mediates genetic expression of IL-6, IL-8, and monocyte chemoattractant protein-1, which are all highly associated with vascular inflammation.[36] It plays a critical role in atherosclerosis in ESRD patients.[37–39] Mallat et al. had found that IL-18 binding protein can reduce atherosclerosis.[40] In our study, the plasma IL-18 level was reduced after six months of HDF, which may result in less IL-18 related atherosclerosis.

Patients dialyzed with HDF have lower serum inflammatory cytokines, which may result from cytokine removal or reduced inflammatory stimulation in HDF. Cytokine removal may result from high-volume convection; reduced inflammation may result from better microbiological control in HDF.[41] Schiffl et al. also found that ultrapure dialysis can lower down the inflammatory stimulation with reduced serum CRP & IL-6 level after 12 months.[42] HEMO study data has shown that higher dialysis doses do not improve survival.[43] However, recent data from DOPPS has shown that high-efficiency HDF reduces mortality risk by 0.65,[44] which may be an effect of cytokine reduction.

CONCLUSION

In this study, plasma TNF-α and IL-18 are decreased while the levels of CRP and IL-6 were unchanged after six months of HDF. The reduction of pro-inflammatory cytokines may result in less vascular inflammation and atherosclerosis; and these may have good effects on survival rates in HDF patients.

DECLARATION OF INTEREST

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.