1. Introduction
Considering that cardiovascular diseases (CVD) are the leading cause of death in children with CKD [
1], assessing cardiovascular risks and detecting early markers of CVD are important. Unlike in adults, congenital anomalies of the kidney and urinary tract (CAKUT) are the most common etiologies of CKD in childhood [
2]. Hypertension has long been a defined risk factor for both CVD and CKD. Over 50% of children with early-stage CKD had abnormalities on 24 h ambulatory blood pressure monitoring (ABPM) in our previous studies [
3,
4]. ABPM is a powerful tool for detecting missed hypertension and is strongly associated with known risk factors for end-stage kidney disease (ESKD) [
5]. In addition, noninvasive assessments of arterial stiffness and left ventricular hypertrophy are also surrogate markers of early events of CVD. These include carotid intimal and medial thickness (cIMT), pulse wave velocity (PWV), ambulatory arterial stiffness index (AASI), left ventricular (LV) mass, and LV mass index (LVMI) [
3,
5,
6]. As CAKUT often develops early in life and may progress to ESKD in childhood or early adulthood, early detection of the abovementioned CVD surrogate markers can help prevent further mortality and morbidity in pediatric CKD.
In addition to functional and structural surrogate markers, biomarkers are also important in early detection of CVD. Proteomics techniques have been used to identify various biomarkers of kidney diseases [
7]. Combined with mass spectrometry, isobaric tags for relative and absolute protein quantification (iTRAQ) have been valuable in discovering potential protein biomarkers [
8]. Our previous study using the iTRAQ technique identified several potential proteins expressed differentially between children with CKD with or without ABPM abnormalities, including complement factor H (CFH), CFH-related protein-2 (CFHR2), and CFH-related protein-3 (CFHR3) [
9].
The complement system is a critical part of innate immunity to protect our body against infection and tumors [
10]. Three main pathways, including the classical pathway, the lectin pathway, and the alternative pathway, contribute to complement activation. CFH is a main regulator of the alternative pathway to inactivate C3b and then to block the amplification loop of C3 convertases and the terminal pathway [
11]. The CFHR proteins can be divided into two groups according to dimerization or not, as circulating in plasma: CFHR2 existing in dimeric form belongs to group 1, while group 2, including CFHR3, is oligomeric [
12]. Considering the competition between CFHRs and CFH for binding to C3b, as described in [
9], we assumed that simultaneous analysis of CFH, CFHR2, and CFHR3 might provide more information to reflect the homeostasis of the complement system.
In this study, we aimed to discover the association between CFH, CFHR2, CFHR3, and ABPM anomalies, parameters of cardiosonography, and arterial stiffness in children with CKD.
2. Materials and Methods
The protocol for this Precision Medicine Project of Pediatric Chronic Kidney Disease (PMP-PCKD) study was obtained with consent from the Institution Review Board and Ethics Committee of Chang Gung Medical Foundation, Taipei, Taiwan (201601181A3 and 201701735A3C501). Chronic kidney disease stage was defined by the KDIGO guidelines [
13]. The Schwartz formula was used to calculate the estimated glomerular filtration rate (eGFR) [
14]. The following were the inclusion criteria of the study: CKD patients aged 6–18 years. Exclusion criteria were patients who were pregnant, had CKD stage 5, had ESKD, were receiving a renal transplant, had congestive heart failure, or were unable to receive cardiovascular assessment. All of the participants, including their parents, provided signed consent to participate in the study. As the most common underlying disease in pediatric CKD is CAKUT [
2], the etiologies of CKD were divided into two categories: CAKUT or non-CAKUT.
All participants received cardiovascular assessments, a fasting blood draw, and spot urine collection. Urine was stored in a −80 °C fridge until examination. The hospital central lab was responsible for measuring blood urea nitrogen (BUN), creatinine (Cr), uric acid, glucose, hemoglobin (Hb), hematocrit (A), sodium (Na), potassium (K), calcium (Ca), phosphate (P), urine total protein-to-Cr ratio (UPCR), triglyceride (TG), and low-density lipoprotein (LDL). Plasma CFH was examined using an ELISA kit (catalog number: CSB-E08931h, CUSABIO TECHNOLOGY LLC, Houston, TX, USA), CFHR2 was examined using an ELISA kit (catalog number: CSB-EL005275HU, CUSABIO TECHNOLOGY LLC, Houston, TX, USA), and CFHR3 was examined using an ELISA kit (catalog number: CSB-EL005276HU, CUSABIO TECHNOLOGY LLC, Houston, TX, USA). All three kits’ inter-assay coefficients of variations were <10%.
We measured office blood pressure and ABPM for all subjects in this study. Before measuring office BP, children rested for 5 min. Then, we recorded the seated non-dominant arm BP three times at one-minute intervals. ABPM was conducted on an Oscar II monitoring machine (SunTech Medical, Morrisville, NC, USA) for 24 h. The ABPM was equipped by a trained nurse as previously described [
15]. The BP was measured every 20 min from 7 a.m. to 10 p.m. and every 30 min from 10 p.m. to 7 a.m. Moreover, the patients recorded the timing of waking, sleeping, and any activities that would fluctuate BP. Abnormal ABPM profiles were determined as follows: (1) daytime BP, nighttime BP, systolic BP (SBP), or diastolic BP (DBP) ≥ 95th percentile basis as per the ABPM reference [
16]; (2) BP load ≥ 25% in daytime or nighttime SBP or DBP; (3) BP load decrease < 10% during sleep compared to the average awake BP load. In addition, the AASI (ambulatory arterial stiffness index) was counted by the formula of 1-DBP/SBP [
13]. Cardiosonography was operated by a pediatric cardiologist with ultrasound (Philips, Bothell, WA, USA). The LV mass was calculated in the M-mode setting. The LVMI (left ventricular mass index) was also counted by dividing LV mass to height
3 [
17]. The PWV (pulse wave velocity), a marker of arterial stiffness, was examined by a ProSound7 ultrasound (ProSound7, e-tracking system; Aloka Co., Tokyo, Japan).
Categorical variables are described as numbers and percentages. Continuous variables are presented as medians and interquartile ranges (IQRs; 25th−75th percentile). We used the Mann–Whitney U-test or chi-square test to test the variables’ difference between two groups and Spearman’s rank correlation coefficient to identify the associations of variables. A linear and logistic regression model were also operated, followed by entering multivariable analyses for adjusting related parameters for CFH, CFHR2, CFHR3, and CV risk markers. A p < 0.05 was considered statistically significant. All analyses were conducted on Statistical Package for the Social Sciences (SPSS) software 22.0 (Chicago, IL, USA).
4. Discussion
The major findings of this study were as follows: (1) the use of office BP alone would have missed the detection of hypertension in more than half the children; (2) children with CKD with non-CAKUT etiology had a higher plasma CFHR3 level compared to the CAKUT group; (3) CFHR3 was negatively correlated with the LV mass and LVMI in the non-CAKUT group, whereas CFH was positively correlated with the LV mass and LVMI in the CAKUT group; (4) CFHR2 was associated with 24 h hypertension and nighttime hypertension detected by ABPM in children with CKD; and (5) a lower plasma CFHR3 level was noted in the non-CAKUT group with daytime hypertension and an increased BP load. We present a brief summary of their relationships in
Figure 2.
In the current study, the non-CAKUT group had a higher eGFR and UPCR. As the normal eGFR number is defined as more than 90 mL/min/1.73 m
2 [
13], both groups actually displayed normal renal function. Most non-CAKUT children had glomerular diseases associated with proteinuria, which is caused by increased glomerular permeability. Our results in the non-CAKUT group were consistent with previous studies showing a higher UPCR than children with CAKUT [
2,
4].
We also noted that there is lack of correlation between CFH proteins and eGFR and UPCR. Although a previous study revealed that urinary CFH protein excretion correlated negatively with eGFR [
18], whether urinary excretion of CFH protein increases with declining GFR due to increased glomerular permeability or impaired tubular reabsorption remains unknown. In addition, proteinuria is a well-known marker for progression of CKD, independent of the underlying kidney disease. Though increasing evidence supports the role of complement in proteinuria-mediated kidney damage [
19], plasma CFH protein level is not correlated with UPCR or eGFR in this study, most likely because most children with CKD are still in the early stages.
Prior research has demonstrated that ABPM is superior to office BP measurement in its ability to distinguish patients at high risk for target-organ damage and has a strong relationship to LVM [
20]. Although our results revealed a high prevalence of BP abnormalities of ABPM, there was no difference regarding an abnormal ABPM profile between the CAKUT and non-CAKUT group. Currently no studies report the association between CFH and related proteins with cardiovascular risk in pediatric CKD. To the best of our knowledge, we are the first to report the impact of CFH and related proteins on the CVD risk in children with CKD.
CFH plays a key role as a regulator in the alternative pathway of the complement system. CFH participates in inflammation and is associated with the pathogenesis of many diseases, including CVD [
21,
22]. Previous studies have shown an association between CFH genetic polymorphism and hypertension [
23], indicating that CFH may be a significant predictor of CVD [
24]. In support of the notion that CFH is related to CVD risk, our results demonstrated that CFH had a positive correlation with the LV mass and LVMI in children with CKD.
The precise role of CFHR proteins is still unclear as to the inhibitory effect thought initially, which is in contrast to recent studies reporting the enhancing activation of the complement pathway [
11]. In our data, CFHR2 had a positive correlation with the LV mass, 24 h hypertension, and nighttime hypertension in children with CKD. In the CAKUT group, CFHR2 was also associated with an increased BP load. CFHR2 is considered an inhibitory regulator of the C3 alternative pathway convertase, acting simultaneously with CFH with a synergic suppression effect [
25]. Another study revealed that a higher urine CFHR2 level was associated with a greater risk of death in patients with proteinuric diabetic kidney disease [
26]. As shown in
Figure 1, we, therefore, hypothesized that CFHR2 may play a role in the pathogenesis of CVD in pediatric CKD and may serve as a predictor of early CVD risk, especially in the CAKUT group.
Regarding CFHR3, it presents homologous domains of CFH, sharing some binding characteristics of CFH and, thus, was considered to have CFH cofactor activity [
11]. The circulating CFHR3 level was greatly influenced by the gene deletion of the CFHR3-CFHR1 allele and gene variants of the CFHR3 gene [
10]. Polymorphism of the CFHR3 gene is associated with an increased risk of atypical hemolytic uremic syndrome [
10,
27]. In line with previous studies showing that a higher CFHR3 level in lupus nephritis was positively correlated with SLEDAI values [
28], our results indicated that the CFHR3 level was higher in the non-CAKUT group.
Interestingly, our study revealed a negative correlation of CFHR3 with the LV mass, LVMI, daytime hypertension, and increased BP load in children in the non-CAKUT group. Whether high CFHR3 reduces CV risks or CVD induces compensatory increases in CFHR3 deserves further elucidation. Additionally, future studies are required to elucidate how CFHR3 works differentially in CAKUT and non-CAKUT CKD and to determine their underlying mechanisms. Our data support the notion that early evaluation of CFH, CHHR2, and CFHR3 in children with CKD will have clinical utility in discriminating CV risk. Although the proportion of abnormal BP load and CV risk markers did not differ between the CAKUT and non-CAKUT group, CFH-related markers may differentially predict CV risks in these two groups that have somewhat different etiologies.
Our study had some limitations. Firstly, we only studied CFH, CFHR2, and CFHR3 levels at one point of time, although there might be serial changes of these proteins at different time points, and there are other proteins in the CFH-related protein family. As immunosuppressive therapy could affect these proteins [
29], whether serial analyses of CFH-related proteins are helpful in monitoring disease activity or therapeutic response deserves further elucidation. Considering different CFHRs can compete with CFH for binding, further research is warranted to investigate the individual role of the CFH-related protein family in children with CKD. Secondly, we did not determine the levels of CFH, CFHR2, and CFHR3 in the normal child population. Lack of normal reference and the pathologic cut point value of the proteins may limit further clinical translation. Thirdly, our data were from a single-centered cohort study; whether children with CKD from one hospital can represent the whole pediatric CKD population remains uncertain. Further multicenter studies are needed to support our findings. Last, the group of diseases encompassed by non-CAKUT did not have a single disease mechanism. Although complement-mediated kidney diseases belong to the non-CAKUT category [
30], our results indicate that plasma CFH-related protein levels vary in different etiologies of the non-CAKUT group. Thus, further subgroups in larger cohorts by different etiologies in the non-CAKUT category will improve the prevention and treatment of CKD, with the aid of CFH-related proteins.