European Journal of Cardiovascular Medicine

Patients with end-stage kidney disease (ESKD), especially those on dialysis are burdened with a high cardiovascular (CV) morbidity and mortality. Compared to the general paediatric population, children with end-stage kidney disease (ESKD) have an increased risk for death, particularly from cardiovascular disease (CVD) due to the development of accelerated arteriosclerosis and premature cardiomyopathy ^1–4. Cardiovascular (CV) causes accounted for 41% of all deaths in children with ESRD in the paediatric Dutch study and for 21.3% in the recent NAPRTCS report ^2,5. The annual report of USRDS indicated that CV mortality among children with dialysis rose from 17.7 deaths/1000 patient-years in 1991 to 23.4 in 2005, which points out the emerging importance of CVD in uraemic children ⁶. Although, this particular preponderance has traditionally been associated with numerous traditional CV risk factors, much recent interest has focused on non-traditional risk factors such as inflammation and endothelial dysfunction ^1,7,8. There is some evidence in adults that haemodialysis may be associated with higher inflammatory mileu than peritoneal dialysis ⁹ but no such studies are available in children. This may be because haemodialysis involves exposure to extracorporeal circuit and also because of the intermittent nature of the procedure.  Several factors can be responsible for the haemodialysis-associated inflammatory response like endotoxin exposure, dialyser membrane bio-incompatability and muscle and body protein breakdown. Peritoneal dialysis is a daily and more continuous process with less of extracorporeal circuit exposure.  At present, there is no clear evidence to suggest any significant difference in the systemic inflammatory burden based on the type of dialysis modality received (i.e., haemodialysis versus peritoneal dialysis) ¹⁰. High-sensitive C‑reactive protein (hsCRP), interleukin‑6 (IL-6) and Fibroblast growth factor 23 (FGF23) are important serum markers of inflammation. The current study measured hsCRP, IL‑6 and FGF23 levels in children with end stage kidney disease undergoing dialysis and compared the levels in those undergoing haemodialysis and continuous ambulatory peritoneal dialysis (CAPD).

This was a prospective cross sectional hospital based study involving children 1-18 years of age with ESKD on maintenance HD or PD for at least 3 months at SAT hospital, Government medical college, Thiruvananthapuram. Children with systemic illnesses, signs of liver diseases, diabetes, thyroid dysfunction or catheter related blood stream infection or PD peritonitis or any other infection within the last 4 weeks were excluded from the study. Study was done with clearance of institutional ethics committee and institutional review board.

For children on HD, blood samples were taken from the venous limb of the HD catheter pre-dialysis. For children on PD, blood samples were drawn from the peripheral vein. Blood samples were drawn in the middle of the week for all patients. Blood was allowed to clot for 10-20 minutes at room temperature and centrifuged at 2000 -3000 RPM for 20 minutes before serum was separated. Serum samples were stored at -20°C for up to 1 month. When sediments occurred during storage, centrifugation was performed again. Serum sample and reagents were brought to room temperature prior to performing the assay. IL-6, hs CRP and FGF 23 levels were estimated using ELISA kits from ImmunoTag, Genotechnology Inc.

hsCRP ELISA test: Range: 0.05mg/L-12mg/L, Sensitivity: 0.01mg/L. Principle -The plate was pre-coated with Human HS-CRP antibody. When serum sample was added, HS-CRP present in the sample binds to antibodies coated on the wells. Then biotinylated Human HS-CRP Antibody was added which binds to HS-CRP in the sample. Then Streptavidin-HRP was added which bound to the Biotinylated HS-CRP antibody. After incubation, unbound Streptavidin-HRP was washed away during a washing step. Substrate solution was then added and color developed in proportion to the amount of Human HS-CRP. The reaction was terminated by addition of acidic stop solution and absorbance was measured at 450 nm.

Interleukin 6 Receptor ELISA test: Range: 0.1pg/ml – 40pg/ml, Sensitivity: 0.042pg/ml. Principle - The plate was pre-coated with human IL6R antibody. When serum sample was added, IL6R present in the sample binds to antibodies coated on the wells. Then biotinylated human IL6R Antibody was added and binds to IL6R in the sample. Then Streptavidin-HRP was added and binds to the Biotinylated IL6R antibody. After incubation, unbound Streptavidin-HRP was washed away during a washing step. Substrate solution was then added and color developed in proportion to the amount of human IL6R.  The reaction was terminated by addition of acidic stop solution and absorbance was measured at 450 nm.

Fibroblast Growth Factor-23 ELISA test: Range: 5pg/ml - 1500pg/ml. Sensitivity: 2.49pg/ml. Principle - FGF-23 present in the sample binds to antibodies coated on the wells. Then biotinylated human FGF-23 Antibody was added and binds to FGF -23 in the sample. Then Streptavidin-HRP was added and binds to the Biotinylated FGF-23 antibody. After incubation, unbound Streptavidin-HRP was washed away during a washing step. Substrate solution was then added and color develops in proportion to the amount of human FGF-23. The reaction was terminated by addition of acidic stop solution and absorbance was measured at 450 nm.

The inflammatory marker levels were compared between the two groups as well as correlated with clinical and laboratory parameters.

Statistical Analysis -All continuous variables were expressed as mean and standard deviation and categorical variables expressed as numbers (percentages). Student's t test, mann-whitney u test, and fisher's exact test were performed to determine differences between groups. The binomial test or χ² test was used for comparison of categorical data. Pearson's or spearman's correlation was applied to determine the relationship and the strength of the association between the variables in bivariate correlation. A probability value of p < 0.05 was considered statistically significant.

A total of 16 children were included in the study (10 on PD and 6 on HD, age range 2.98 - 18.30 years; average age -12.18±4.79 years, 62.50% males.). All the patients on PD were on Continuous Ambulatory PD (CAPD). The study subjects were on a dialysis modality for a mean of 17.51 months (Range – 4.5-79 months). Baseline characteristics of the study population is depicted in table 1. The two groups were different in age (14.95±3.08 years in HD group vs 7.57±3.32 years in PD group, p<0.001). Other parameters like sex, presence of hypertension, anemia, residual urine output, left ventricular hypertrophy, dialysis vintage and erythropoietin (EPO) dose were not significantly different between the two dialysis groups. Among the baseline laboratory parameters, serum cholesterol and calcium level was observed to be higher in PD group compared to HD group. Comparison of baseline clinical and laboratory markers based on the method of dialysis is shown in Table 2.

Inflammatory marker levels - The mean hsCRP was 3.20±0.69 mg/L in dialysis patients (Range: 2.46 – 4.41 mg/L). The mean IL-6 was 1.69±0.46 pg/ml (Range: 0.57- 2.27 pg/ml). The mean FGF23 was 811.54±128.6 pg/ml (Range: 472.12-982.73 pg/ml).

All the inflammatory markers - hsCRP, FGF 23 and IL6 were not significantly different between the dialysis groups (p>0.05). The mean hsCRP was 3.14±0.59 mg/L in HD group and 3.28±0.87 mg/L in PD group (p=0.515). The mean IL-6 was 1.81±0.37 pg/ml in HD group and 1.50±0.56 pg/ml in PD group (p=0.913). The mean FGF23 was 821.7±137.9 pg/ml in HD group and 794.6±121.6 pg/ml in PD group (p=0.233).  

Correlation of inflammatory parameters - Among the inflammatory markers, FGF23 showed significant moderate positive correlation with dialysis vintage (R=0.509, p=0.044). Except FGF23 and dialysis vintage, all other parameters did not show significant correlation even though they showed mild positive or negative trends (Table 3). Assessment of relationship between inflammatory markers showed that there is moderate positive correlation between FGF23 and IL6 which was statistically significant (R=0.547, p=0.028). There was no significant correlation of hsCRP with FGF-23 or IL6 (Table 4).

Table 1: Distribution of baseline clinical and laboratory parameters of the study population

Table 2: Comparison of baseline clinical and laboratory parameters between HD and PD groups

a: Independent T test, b: Fisher’s exact test, c: Mann Whitney U test, ***p<0.001, **p<0.05

Table 3: Correlation of inflammatory parameters with baseline clinical and laboratory parameters

a: Spearman Rank Correlation, b: Point biserial correlation, *p<0.05

Table 4: Correlation between inflammatory markers

  Spearman Rank Correlation, *p<0.05

The PD group was younger compared to HD group. The mean age of haemodialysis group was 14.95±3.08 years and that of peritoneal dialysis group was 7.57±3.32 years (p<0.001). This is mainly because younger children are prone to be put on PD due to issues with vascular access.

The mean hsCRP was 3.20±0.69 mg/L in dialysis patients (Range: 2.46 – 4.41 mg/L). AHA classifies CVD risk groups based on hsCRP concentration: low risk, <1.0 mg/L; moderate risk, 0.1–0.3 mg/L; and high risk, >3 mg/L. Thus our patients fall under high risk group for CVD.

The mean FGF-23 was 811.54±128.6 ng/ml (Range: 472.12-982.73 ng/ml). This was much higher than the normal FGF23 levels reported in children 1-18 years ie <=57pg/ml ¹¹. However, the mean IL-6 was only 1.69±0.46 pg/ml (Range: 0.57- 2.27 pg/ml), which was within normal range. An IL-6 level of less than 5pg/ml is considered normal. Thus out of the three markers, hsCRP and FGF23 were very much increased whereas IL-6 levels were normal in our study. In a study by Elshamaa et al, hsCRP were determined in 30 paediatric patients on HD and in 20 healthy controls. The hsCRP values were significantly higher in HD patients than in healthy controls ¹². The percentage of HD paediatric patients with CRP >10 mg/l was 30%. Similarly FGF 23 levels are known to be very much raised in patients on dialysis. In a study by Torguet-Escuader, FGF 23 levels increased from 31.32± 12.46ng/ml to 702±436.4ng/ml as the patients progressed from stage 1 CKD to RRT ¹³.

There are several factors that can potentially cause differences in inflammation between the dialysis modalities. In HD, dialyzer membranes are thought to affect plasma cytokine concentrations, in part by activating leukocytes. The dialysate is also a potential source of cytokine induction. Infections at exit sites or in catheter tunnels can cause inflammation in PD. When we compared the inflammatory markers between patients on HD and PD, we could not observe any significant difference in the levels of these markers. Data on inflammation in HD versus PD have been conflicting. Two studies concluded that there is no difference in IL-6 and CRP in HD compared to PD patients ^14,15 whereas Haubitz et al. showed an increase in CRP after starting HD but not PD; and the PD group had CRP levels similar to healthy controls ¹⁶. In another study involving 228 prevalent HD patients  and 80 prevalent PD patients prevalence of inflammation was high in both groups of dialysis patients, two markers of inflammation, IL-6 and hs-CRP, as well as albumin and the cardiac biomarker N-terminal pro-B-type natriuretic peptide (NT-proBNP), were all higher in HD than in PD patients ¹⁷. But in this study the PD patients had a shorter dialysis vintage where as in our study there was no difference in dialysis vintage between HD and PD groups. However FGF23 level showed significant moderate positive correlation with dialysis vintage in our study.

Assessment of relationship between inflammatory markers showed that there is moderate positive correlation between FGF23 and IL6 which was statistically significant but there was no significant correlation of hsCRP with FGF-23 or IL6. Snaedal et al have shown positive correlations between hs-CRP and IL-6 in both HD and PD cohorts ¹⁷. The association between levels of FGF23 and the inflammatory markers IL-6, C-reactive protein (CRP), TNF-α, and fibrinogen was assessed in a cross-sectional analysis of 3879 participants enrolled in the Chronic Renal Insufficiency Cohort (CRIC) study between June 2003 and September 2008. FGF23 correlated directly with IL-6 (r=0.4), CRP (r=0.2), TNF-α (r=0.4), and fibrinogen (r=0.3; P<0.001 for each). In univariate and multivariable-adjusted linear regression analyses, natural log (ln) transformed FGF23 was significantly associated with lnIL-6, lnCRP, lnTNF-α, and fibrinogen (P<0.001 for each) ¹⁸. 

Children on maintenance dialysis have high inflammatory load as evidenced by high hsCRP and FGF-23 levels. Based on the hsCRP level, they fall into high cardiovascular risk category which have implications on morbidity and mortality. We could not observe any difference in the inflammatory marker levels between HD and PD patients. Greater dialysis vintage may be associated with higher FGF23 levels and thus higher inflammation.

Strengths and limitations of the study.

Very few studies on inflammatory load in paediatric dialysis are available in the literature. The children included in this study were on dialysis for a mean of 17.51 months, which is a significant duration to elicit any inflammatory response. The groups on HD and PD were well matched regarding baseline characteristics except for lower age of PD subjects. The major limitation of the study is the limited sample size which might have restricted the study’s ability to detect difference between HD and PD groups.

Acknowledgement

State board of medical research for funding the study

1. Mitsnefes MM. CV complications of paediatric chronic kidney disease, Pediatr Nephrol, 2008, vol.23 (pg.27-39)
2. Groothoff JW, Gruppen MP,  Offringa M, et al. Mortality and causes of death of end-stage renal disease in children: a Dutch cohort study, Kidney Int, 2002, vol. 61(pg. 621-629)
3. Parekh RS,  Carrol CE,  Wolfe RA, et al. CV mortality in children and young adults with end-stage kidney disease,  J Pediatr, 2002, vol. 141(pg. 191-197)
4. Chavers BM, Li S, Collins AJ, et al. CV disease in paediatric chronic dialysis patients,  Kidney Int, 2002, vol. 62 (pg. 648-653
5. North American Paediatric Renal Trials and Collaborative Studies (NAPRTCS) Annual Report 2008 https://web.emmes.com/study/ped/annlrept
6. United States Renal Data System System, 2007 Annual Report. http://www.usrds.org/adr_2007.htm
7. Kari JA,  Donald AE, Vallance DT, et al. Physiology and biochemistry of endothelial function in children with chronic renal failure, Kidney Int, 1997, vol. 52(pg. 468-472)
8. Mitsnefes MM. Paediatric end-stage renal disease: heart as a target, J Pediatr, 2002, vol. 141 (pg. 162-164)
9. Kenneth Yong, Gursharan Dogra , Neil Boudville , Wai Lim . Increased Inflammatory Response in Association with the Initiation of Haemodialysis Compared with Peritoneal Dialysis in a Prospective Study of End-Stage Kidney Disease Patients. Perit Dial Int. Jan-Feb 2018;38(1):18-23.
10. Zoccali, G. Tripepi, and F. Mallamaci, “Dissecting inflammation in ESRD: do cytokines and C-reactive protein have a complementary prognostic value for mortality in dialysis patients?” Journal of the American Society of Nephrology, vol. 17, no. 12, supplement 3, pp. S169–S173, 2006.
11. Stanczyk M, Chrul S, Wyka K, Tkaczyk M. Serum intact fibroblast growth factor 23 in healthy paediatric population. Open Med (Wars), 2021;16(1):1022-1027
12. Elshamaa MF, Sabry S, Nabih M, Elghoroury EA, El-Saaid GS, Ismail AA. Oxidative stress markers and C-reactive protein in paediatric patients on haemodialysis. Ann Nutr Metab. 2009;55(4):309-16.
13. Pere Torguet-Escuder, Bernat Guasch-Aragay, Jordi Calabia-Martínez, Nàdia Martín-Alemany, Isabel García-Méndez, Gerard Maté-Benito, Esther Clapés-Sánchez, Mònica Sabater-Masdeu, José M. Fernández-Real, Martí Vallès-Prats. Fibroblast growth factor 23 (FGF 23) and phosphocalcic metabolism in chronic kidney disease. Nefrologia 2012;32(5):647-54.
14. Filiopoulos V,  Hadjiyannakos  D,  Takouli  L,  Metaxaki  P,  Sideris  V,  Vlassopoulos  Inflammation  and  oxidative  stress  in  end-stage  renal  disease  patients  treated  with haemodialysis  or  peritoneal  dialysis.  The  International  journal  of  artificial  organs 2009;32:872-82.
15. Borazan A, Ustun H, Ustundag Y, et al. The effects of peritoneal dialysis and haemodialysis on serum  tumor  necrosis  factor-alpha,  interleukin-6,  interleukin-10  and  C-reactive- protein levels. Mediators of inflammation 2004;13:201-4.
16. Haubitz M, Brunkhorst R, Wrenger E, Froese P, Schulze M, Koch KM. Chronic induction of C-reactive protein by haemodialysis, but not by peritoneal dialysis therapy. Peritoneal dialysis international  :  journal  of  the  International  Society  for  Peritoneal  Dialysis 1996;16:158-62.
17. Sunna Snaedal, Abdul R. Qureshi, Sigrún H. Lund, Guna Germanis, Britta Hylander, Olof Heimbürger, Juan   Carrero,  Peter  Stenvinkel   and  Peter  Bárány. Dialysis  modality  and nutritional  status  are  associated  with variability  of  inflammatory  markers. Nephrol  Dial  Transplant  (2016)  31:  1320–1327.
18. Jair Munoz Mendoza, Tamara Isakova, Ana C. Ricardo, Huiliang Xie, Sankar D. Navaneethan, Amanda H. Anderson, Lydia A. Bazzano, Dawei Xie, Matthias Kretzler, Lisa Nessel,  Lee Hamm, Lavinia Negrea, Mary B. Leonard, Dominic Raj, Myles Wolf,and for the Chronic Renal Insufficiency Cohort. Fibroblast Growth Factor 23 and Inflammation in CKD. Clin J Am Soc Nephrol. 2012 Jul; 7(7): 1155–1162.