DOI: https://doi.org/10.21203/rs.3.rs-2332044/v1
Background: Numerous equations have been reported for the estimation of glomerular filtration rate (eGFR) based on serum creatinine. Serum cystatin C-based equation has been recently demonstrated to be better in estimating GFR. This study was conducted to examine the agreement between eGFR measured by equations based on serum creatinine (eGFRCr) and cystatin C (eGFRCys) or both (eGFRCrCys) in children with chronic kidney disease (CKD).
Methods: This prospective observational study was conducted in children less than 14-years with CKD stage 2-4. CKiD equation was used for calculation of eGFRCr and eGFRCys. Considering the eGFRCr as reference standard, we assessed the agreement of eGFRCr with eGFRCys and eGFRCrCys estimated by constructing a Bland-Altman plot and visually estimating the distribution of points representing the difference between eGFRCr and eGFRCys or eGFRCrCys against the line of zero difference.
Results: A total of 60 patients (54 boys) with mean age of 88±47 months were enrolled. Overall, 57%, 35% & 8% children had CKD stage 2, 3, and 4 respectively. The mean eGFRCr, eGFRCys and eGFRCrCys was 58 (19), 55 (21) and 62 (12) ml/min/1.73 m2, respectively. Bias between eGFRCr and eGFRCys was 2.8 (95% CI: -1.03 to 6.6) ml/min/1.73 m2. Bias between eGFRCr and eGFRCrCys was -4.5(95% CI: -6.5 to -2.5) ml/min/1.73 m2.
Conclusion: Serum creatinine-based equation slightly overestimates the GFR when compared with eGFRCrCys with overall average agreement between equations in children with pre-dialysis CKD. For GFR estimation, the combination of serum creatinine and serum cystatin C is more precise than either marker alone.
Serum creatinine-based estimation of glomerular filtration rate (eGFR) is the standard of care for the diagnosis and staging of chronic kidney disease (CKD)[1]. Numerous equations have been reported for the estimation of GFR based on serum creatinine[2]. Staging of CKD is essential for planning the management strategy and determining the long-term prognosis[3]. Hence, accurate measurement of eGFR is necessary. It is also important for follow-up and dose modification of medications which are nephrotoxic or are excreted by the kidneys[3].
The gold standard methods of assessing GFR require measurement of an ideal filtration marker[4, 5]. However, they are technically challenging, expensive and time-consuming[6]. Serum creatinine-based estimated GFR equations are widely used in clinical practice; however, serum creatinine may be affected by the intake of protein and muscle mass. Cystatin C has been introduced as an alternative endogenous marker. Cystatin C is less affected by muscle mass and diet than creatinine; hence it was widely anticipated that cystatin C would provide a more accurate estimate of GFR[7]. In addition, it has been proposed that cystatin C is a more sensitive marker for kidney function than serum creatinine in conditions with only a moderate decrease in GFR, hence giving the added advantage in the creatinine blind area of initial renal impairment[8]. Equations combining creatinine and cystatin C provides the most precise and accurate estimates[9]. It may be useful to consider the more widespread use of GFR estimates based on cystatin C, either alone or in combination with creatinine.
Many equations have been developed to estimate GFR which vary in their precision, accuracy and practicality; however, in children, the most widely recognized equation is the updated Bedside Schwartz formula from the CKiD study[7]. The aim of our study was to examine the performance and agreement between eGFR measured by equations based on serum creatinine (eGFRCr) and cystatin C (eGFRCys) or both (eGFRCrCys) in children with CKD.
A prospective observational study was conducted in the Pediatric Nephrology Unit of a tertiary care centre from July 2018 to July 2021. All consecutive children with CKD presenting to the renal children clinic were assessed for eligibility. Children aged 1–14 years and with CKD stages 2–4 were enrolled. Children with a recent history of edema or jaundice and those on dialysis were excluded. Informed written consent was taken from parents/ guardians before the enrolment of the children.
The demographic, clinical and laboratory parameters were recorded in a pre-design proforma. CKD was defined as per Kidney Disease Improving Global Outcomes (KDIGO) criteria[1]. Complete blood count with peripheral blood smear, serum urea, creatinine, calcium and phosphate, iron studies, vitamin D3, intact parathyroid hormone (iPTH), urinalysis and ultrasound scan of the kidney was done in children. Serum creatinine measurement was done on the same day of the sample collection by Jaffe method on Beckman Coulter AU analyser. A blood sample (2ml) collected was kept at room temperature for 30 minutes before centrifugation for 10 minutes at 2500rpm. Serum was separated and stored at -800 C. Cystatin C was measured within 6 months of collection using Quantikine ELISA Human Cystatin C kit (R&D systems) with Infinite M200PRO ELISA reader, make TECAN. The coefficient of variation ranged from 3.1–6.6%, and the measurement limit ranged from 0.030–0.227 ng/mL. The mean minimal detectable dose was 0.102 ng/mL. CKiD equation was used for the calculation of eGFRCr and eGFRCys[10].
Data were analysed using STATA software. Categorical variables were expressed in terms of frequency. Continuous variables were expressed as mean and standard deviation or median with an interquartile range. Considering the eGFRCr as the reference standard, we assessed the agreement of eGFRCr with eGFRCys and eGFRCrCys estimated by constructing a Bland-Altman plot and visually estimating the distribution of points representing the difference between eGFRCr and eGFRCys or eGFRCrCys against the line of zero difference.
A total of 58 patients (54 boys) with a mean age of 89 ± 48 months were enrolled. Overall, 57%, 35%, & 8% children had CKD stage 2, 3, and 4, respectively. The baseline characteristics of the studied children are reported in Table 1. The mean eGFRCr, eGFRCys and eGFRCrCys was 58 (19), 55 (21) and 62 (12) ml/min/1.73 m2, respectively.
| Variable | Value* |
|---|---|
| Boys n (%) | 52 (90) |
| Age (months) | 89.4 (48) |
| Height (cm) SDS | 113 (25) |
| Weight (kg) SDS | 21.7 (11.7) |
| Serum creatinine (mg/dL) | 0.77 (0.6 to 1.21) |
| Serum Cystatin C (ng/mL) | 1512.98 ± 643.77 |
| Proteinuria (Up/Uc;mg/mg) | 0.3 (0.14 to 1) |
| eGFRCr (ml/min/1.73m2) | 58 (19) |
| eGFRCys (ml/min/1.73m2) | 55 (20.5) |
| eGFRCrCys (ml/min/1.73m2) | 62 (12) |
| *values are presented as median (interquartile range) or mean (standard deviation); SDS: Standard deviation score; Up/Uc: Spot urine protein/ creatinine ratio; eGFRCr & eGFRCys: estimated glomerular filtration rate based on serum creatinine & cystatin C; eGFRCrCys: estimated glomerular filtration rate based on combination of creatinine and cystatin C. | |
GFR estimation based on serum creatinine and cystatin C showed a moderate correlation (r = 0.72; 95% CI: 0.59 to 0.84) as depicted in Fig. 1. Agreement between eGFRCr and eGFRCys was assessed by generating a Bland-Altman plot as depicted in Fig. 2. The limit of agreement between these two methods varied from − 26.4 to 32 ml/min/1.73 m2. Bias between eGFRCr and eGFRCys was 2.8 (95% CI: -1.03 to 6.6) ml/min/1.73 m2. The precision was 19.5 ml/min/1.73 m2.
GFR estimation based on serum creatinine and a combination of cystatin C and creatinine showed a good correlation (r = 0.84; 95% CI: 0.79 to 0.88) as depicted in Fig. 3. Bias between eGFRCr and eGFRCrCys was − 4.5(95% CI: -6.5 to -2.5) ml/min/1.73 m2. Agreement between these two methods was assessed by generating a Bland-Altman plot as depicted in Fig. 4. There was good agreement between eGFRCr and eGFRCys with narrow dispersion of values both above and below the line of zero difference – the limits of agreement varied from − 20 to 11 ml/min/1.73 m2. Precision of the combined equation was 12.4 ml/min/1.73 m2.
In our study, we found that GFR estimation based on an equation using serum creatinine and serum cystatin C showed good concordance and agreement as suggested by a bias of 5.4 (95% CI: -1.03 to 6.6) ml/min/1.73 m2. These findings suggest that there is no systematic deviation between eGFRCr and eGFRCys. Estimation by serum creatinine alone tends to overestimate GFR when compared with the combination of cystatin C and creatinine-based equation. Agreement and precision were better between the eGFRCr and eGFRCrCys equation as compared to eGFRCys and eGFRCr.
Overall, in the last decade, multiple equations have been published to estimate the GFR in the pediatric population[9]. A recent large study which derived an equation for estimation of eGFR without race concluded that an equation using both creatinine and cystatin C is better than an equation using either of them[11]. Similar to this large study, our findings also suggested that using both markers is better for the estimation of kidney function in children with CKD. The study by Hari et al.[12] showed that when compared with measured GFR (estimated by using 99mTcdiethylenetriamine pentaacetic acid (DTPA) scan), cystatin C-based equation tends to overestimate the measured GFR and in contrary, the combination of creatinine and cystatin C based equations underestimate the measured GFR. A study from Saudi Arabia showed that creatinine and combined creatinine and cystatin C-based studies showed good agreement, but alone, the cystatin C-based equation was inconsistent in the estimation of GFR[13]. A study assessing the different GFR estimating equations found that in children with GFR < 60 mL/min/1.73 m2, cystatin C alone based equation was better, but in those with GFR > 60 mL/min/1.73 m2, the equation using a combination of cystatin C and creatinine was better than the equation using either of this biomarker alone[9]. In another study by Chehade et al.[14] evaluating 201 children, they found that an equation combining both creatinine and cystatin C was better than creatinine alone or the CKD-EPI formula. They also showed that the combined equation is better than serum cystatin C alone in a cohort of children chiefly having CKD stages I and II. While most of these studies used a measured GFR for comparison unlike our present study, all studies concluded that using an equation based on the combination of cystatin C and creatinine provided a better estimation of kidney function. The findings from our study also support these observations in children with eGFR between 15 to 90 mL/min/1.73 m2.
One of the limitations of our study was the small sample size across various stages of CKD. Due to limited participants, we could not assess the agreement and bias across various ranges of GFR. Similarly, a limited sample size could have affected the agreement and bias between the various equations for GFR estimation. Another limitation of our study was the lack of the gold standard reference test for measurement of GFR, where we chose to use serum creatinine, the marker which is commonly used in clinical practice as a reference standard test.
Serum creatinine-based equation slightly overestimates the GFR when compared with eGFRCrCys with an overall good agreement between eGFRCr and eGFRCrCys equations in children with pre-dialysis CKD. Findings from this study suggest that a combination of serum creatinine and serum cystatin C-based equation is more accurate than either marker alone for estimating GFR.
Ethical approval: All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee at which the studies were conducted (IEC approval number: INT/IEC/2018/826; dated: 24.05.2018) and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Disclosure of potential conflicts of interest: The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Informed consent: Informed consent was obtained from parents/ guardians of all the individual participants included in the study.
Funding: This work was supported by funding from New Institute Research Grant (L.P. No. 521), Post Graduate Institute of Medical Education and Research, Chandigarh.
Author’s contribution:
LD conceptualized and designed the study, collected clinical data, analysed and interpreted the data, reviewed the literature, and wrote the manuscript. AR performed the laboratory test and interpreted them, supervised overall concept design, critical review of the article for important intellectual content, and final approval of the manuscript. KT supervised overall concept design, critical review of the article for important intellectual content, and final approval of the manuscript. JM analysed and interpreted the data, revised the manuscript for important intellectual content and final approval of the manuscript. All the authors revised and approved the final version of the manuscript.
Acknowledgements: We thank Mr. Ravinder Kumar, Technical Assistant, for his support in performing the laboratory test.
Disclaimers: Nil
Declaration on competing interest: Nil