Applicability of 14 Formulas for Estimating Glomerular Filtration Rate in the Evaluation of Renal Function before and after Nephron-Sparing Surgery in Patients with Renal Tumors

To compare the applicability of 14 equations of estimating glomerular filtration rate (eGFR) before and after nephron-sparing surgery (NSS) for renal function assessment of patients with renal tumors. Preoperative and postoperative GFR is measured by emission computed tomography (ECT) with 99mTc-DTPA as an imaging agent as reference GFR (rGFR) to compare with all formulas. Spearman correlation analysis and Bland–Altman agreement analysis were used to evaluate the correlation between rGFR and eGFR1 to 14 before and after surgery. A total of 50 cases including 22 males and 28 females were included. The results of preoperative eGFR1–14 correlated with rGFR (P < 0.05). The calculation results of all estimation formulas have a significant correlation with preoperative GFR. Preoperative MDRD-I, CKD-EPI SCysC, and FAS Scr-SCysC have good consistency. The CG formula has the highest precision and FAS Scr-SCysC has the highest accuracy. A total of 30 patients followed up after surgery, and postoperative rGFR correlated with CG, CKD-EPI, FAS, and BIS formulas (P < 0.05). But postoperative rGFR has no significant correlation with MDRD and Schwartz (P > 0.05). Postoperative CKD-EPI Scr-SCysC has best consistency, and FAS Scr-SCysC has the highest accuracy and precision. Our data suggest that eGFR equations evaluated by both serum creatinine (Scr) and cystatin C (SCysC) is not necessarily better than those evaluated by one of them alone. Among all enrolled equations, FAS Scr-SCysC is the best one to evaluate postoperative GFR in patients with renal tumors.


Introduction
GFR refers to the number of milliliters of plasma that the kidneys completely filter and clear a certain metabolite in plasma within a unit time [1]. It is an important indicator for effectively evaluating the renal function of patients [2]. Methods based on the clearance of exogenous markers, like inulin, 51Cr-EDTA, and iohexol, are the gold standard for glomerular filtration rate (GFR) measurement, but are cumbersome and infrequently used in clinical settings [3]. At present, the clinically accurate determination of GFR mainly relies on ECT, with proven ease of operation and good repeatability [4]. In clinical work, relying on the biochemical results of renal function in the laboratory, such as Scr, blood urea nitrogen (BUN), and SCysC, calculating GFR through formulas is more convenient, easy, economical, and less harmful to patients [5].
Since the Cockcroft-Gault (CG) formula was widely used in clinical practice in 1976 [6], eGFR formulas based on Scr and SCysC have been developed endlessly.
e National Kidney Foundation (NKF) guidelines recommend the use of the CG formula, MDRD (modification of diet in renal disease) formula, and CKD-EPI (chronic kidney disease epidemiology collaboration) formula [7], but the results of the current studies have not confirmed that the above recommended formula is the optimal one. Studies have shown that Scr-based CKD-EPI has better applicability than MDRD [8,9]. e research results of Liu et al. on adult CKD patients in China show that the CG formula has better applicability [10]. At present, the commonly used formulas in clinical practice are FAS (full age spectrum), MDRD-I, MDRD -II, MDRD-China (MDRD-C), CKD-EPI (Scr), BIS-1 formula, Schwartz's formula, and so on. Estimated glomerular filtration rate (eGFR) is prone to bias in specific situations or special populations [11]. So far, there are few studies in China that systematically evaluate the difference in eGFR calculation results of patients with kidney tumors before and after nephron-sparing surgery. We used all 14 estimating formulas to calculate the GFR of renal tumor patients undergoing nephron-sparing surgery to verify the applicability of various formulas to renal tumor patients in my country.

Sources of Data and Measurements.
A total of 50 patients with renal tumor who underwent NSS in Fujian Provincial Hospital from July 2020 to July 2021 were enrolled. All patients had no renal puncture, interventional surgery, radiotherapy, or chemotherapy, or other antitumor-related treatments before surgery. No history of CKD disease, renal trauma, and surgery. Clinical data include gender, age, height, weight, preoperative creatinine, urea, nitrogen, albumin (Alb), cystatin C, rGFR measured by ECT, and pathological diagnosis, as given in Table 1. e patient's serum creatinine concentration was measured with the alkaline picric acid method using the Beckman CX3 analyzer. Serum cystatin C was measured on the Beckman UniCel DxC800 biochemical analyzer with the latex enhanced immunoturbidimetric method.
99mTc-DTPA imaging determination rGFR: before the examination, the patient drinks 300-500 ml of water, empties the bladder, and takes the supine position. e probe is placed on the back waist, and the field of view includes both the kidneys and bladder. e imaging agent is injected into the vein of one side of the elbow in the form of a "bullet," and two sets of images of blood perfusion and renal function are dynamically collected in two phases. e acquisition field includes the kidneys, ureters, and bladder. e first phase is 1 s/frame, and 60 frames are collected. e second phase is 20 s/frame, and the collection is 29 minutes. After the collection is completed, the empty syringe is collected again and counted for 6 s. Calculate the dose of drug injected into the body according to the count value of the syringe before and after. Use the ROI technology to outline the outline and background area of the kidneys. en, use the supporting software to process the image and generate the kidney map to obtain the GFR of the kidneys.

Statistical Analysis.
e measured height, weight, Scr, BUN, Alb, and SCysC values were, respectively, substituted into the 14 estimation equations included in this study, and eGFR1-14 were calculated, as given in Table 2. Using SPSS 16.0 and MedCalc 15.2.2 statistical software, perform analysis. Spearman correlation analysis and Bland-Altman agreement analysis were used to evaluate the correlation between rGFR and eGFR1-14 before and after surgery.
Accuracy means the degree to which the measured value of the formula is consistent with rGFR, indicating the correctness of the calculation result. e greater the number of close agreement between the eGFR results and rGFR values of multiple patients calculated by each formula, the higher the precision of the results, showing the repeatability and stability of the formula.

Results
A total of 50 cases including 22 males and 28 females, aged 17−80 years old, with an average of 53. 8

Consistency Analysis of Preoperative eGFR and rGFR.
e mean line of eGFR2 (MDRD-I) is closest to 0, indicating that its calculation result is closest to rGFR, as shown in Figure 1. e closest moving average value to 0 is −0.13 (B). Only one calculated eGFR value was outside the 95% CI (B, G, L). e absolute minimum value of the intercept width is −0.2096 (A). e absolute value of the slope is the smallest, which is −0.0144 (L). e absolute value of the slope is the largest at −0.7026 (M).
Among the 14 formulas, the absolute value of the intercept width of eGFR1 (CG formula) is the smallest, which is −0.2096, indicating that eGFR1 (CG formula) has highest accuracy than other formulas. Followed by eGFR12 (FAS Scr-SCysC ) and eGFR7 (CKD-EPI SCysC ) are −5.0456 and −5.4934 respectively. e minimum absolute value of the slope of eGFR12 (FAS Scr-SCysC ) is −0.0144, indicating the highest accuracy. Followed by eGFR1 (CG formula) and eGFR7 (CKD-EPI SCysC ), they are −0.0516 and −0.0592, respectively. Among the CKD-EPI related formulas (eGFR6, eGFR7, and eGFR8), the cystatin C-based formula has the highest 2 Contrast Media & Molecular Imaging CKD-EPI Scr [16]:  consistency than the equation based on creatinine and the combination formula. e FAS Scr-SCysC formula has the highest consistency than those based on creatinine or cystatin C alone.
All calculated GFR values of eGFR8 (CKD-EPI Scr-SCysC ) are within 95% CI means it has better consistency. Except for eGFR4 (MDRD-C) and eGFR10 (FAS Scr ), which have two GFR values outside the 95% CI, there is only one other formula. Among the 14 formulas, the absolute value of the slope and intercept of eGFR12 (FAS Scr-SCysC ) is the smallest, indicating that its accuracy and precision are the highest. e accuracy of eGFR4 (MDRD-C), eGFR7 (CKD-EPI SCysC ), eGFR10 (FAS Scr ), and eGFR1 (CG formula) is higher, for the absolute value of their slopes are all less than 0.1. e absolute values of the intercepts of eGFR10 (FAS Scr ) and eGFR7 (CKD-EPI SCysC ) are smaller than those of other formulas, indicating that the accuracy of these two formulas is higher than others, as given in Table 4.

Discussion
In recent years, with the development of early screening for malignant tumors and the improvement of people's health awareness, the early diagnosis rate of renal tumors has increased significantly. e correct assessment of renal function is of great importance to decide surgical strategy. GFR is a direct indicator for evaluating renal function. GFR measurement of the patients with renal tumor before and after nephron-sparing surgery may be important in treatment strategy decision and prognosis evaluation. e removal rate of inulin is an ideal method for measuring GFR, but procurement of inulin and the cumbersome procedure pose a challenge for its routine clinical use. Among many GFR estimation formulas, how to choose an appropriate formula is a puzzled problem faced by clinicians. In recent years, most of the eGFR evaluation equations have been verified by patients with chronic kidney disease, and the eGFR formula is rarely used in the evaluation of renal function in patients with renal tumors. We use the 99mTc-DTPA renal function imaging method to measure GFR as the reference standard (rGFR) [22] to assess the applicability of 14 commonly used formulas in patients with renal tumors.
e results of the study showed that there was a significant correlation between the GFR calculated by the 14 estimation formulas and rGFR before the surgery. Studies have shown that the combined formula of serum creatinine and cystatin C is better than using one of them alone [23,24]. Among the 14 formulas included in this study, the calculation results of the three formulas including MDRD-I, CG formula, CKD-EPI SCysC , and FAS Scr-SCysC are in good agreement with rGFR. Among them, the CG formula has the highest precision, and FAS Scr-SCysC has the highest accuracy. Renal tumors can occur in almost all age groups [25]. e FAS formula is suitable for patients of all ages and has good applicability for preoperative renal function assessment [19].
In the results calculated after the operation, 5 formulas had no significant correlation (P > 0.05), including MDRD-I, MDRD-II, MDRD-C, MDRD, and Schwartz Scr-SCysC . ese formulas have limitations for the evaluation of renal function in patients with renal tumors after surgery. e Schwartz formula is mainly used to evaluate the renal function of children with CKD [18]. e youngest age of the patients in this study is 17 years old. erefore, it is verified that the Schwartz formula is not highly applicable to patients over 14 years of age. In the evaluation of renal function in adult patients with renal tumors after surgery, this method is not recommended.
In 1999, the US MDRD research team found that age, gender, and race were independent variables of GFR and derived the original MDRD formula [15]. In 2000 and 2007, MDRD-I and MDRD-II were obtained by improving the formula [12,13]. In 2006, the Chinese eGFR project  Contrast Media & Molecular Imaging 2020 to evaluate the eGFR estimation formula [26]. On November 4, 2021, NEJM officially published the results of two research teams on eGFR estimation of race coefficient. e results of these two studies show that the eGFR estimation formula based on serum cystatin C is more accurate than the estimation formula based on serum creatinine [27]. So, they recommended the following: the race coefficient has limited influence on the eGFR calculation, and the race coefficient can be removed; each clinical laboratory should try its best to detect serum cystatin C; and further research to find better markers for eGFR estimation [26,27]. e population included in the MDRD study is mainly concentrated in patients aged 18-70 [15]. Past studies have shown that when patients have low GFR, the MDRD simplified formula is less accurate and tends to underestimate renal function [28]. e renal function of patients after renal surgery has declined to a certain extent. erefore, the MDRD formula is not suitable for postoperative renal function assessment. But, we should not ignore the possibility of errors caused by the reduction in sample size.
In 2016, Pottel et al. proposed an eGFR equation based on Scr/Q Scr that is applicable to all-year-olds (full age spectrum, FAS), where Q Scr is the median number of Scr for healthy people of a specific age and gender [19]. In 2017, the research group used cystatin C instead of Scr and proposed the FAS SCysC formula and the FAS Scr-SCysC formula [20]. Studies on the applicability of the FAS formula in the Chinese population have confirmed that FAS Scr-SCysC has higher applicability in China [29]. In this study, GFR calculated by FAS Scr-SCysC formula after surgery has the highest accuracy and precision. Patients after surgery are recommended to use this equation for renal function assessment.
In 2009, the United States Chronic Kidney Disease Epidemiology Cooperative Group proposed the CKD-EPI Scr formula using Scr, age, gender, and race as calculation variables [16]. In 2012, the CKD-EPI research group proposed the CKD-EPI SCysC formula and the CKD-EPI Scr-SCysC formula based on standardized Scr and SCysC [17]. In the comparison and verification of the three CKD-EPI formulas, it is found that the combined formula has no significant difference in calculation deviation compared with the two separate formulas, but the precision and diagnostic accuracy are higher [30]. Our results show that the calculation results of the CKD-EPI Scr-SCysC formula are highly consistent, and the postoperative GFR calculated by CKD-EPI SCysC is more accurate than the other two formulas. ese reflect the advantage of cystatin C in early recognition of renal function damage. CKD-EPI formula is mainly derived and established based on CKD patients as the baseline data. e applicability of the formula is different because of the significant differences in renal pathology between patients with renal cancer and CKD. e BIS-1-2 formula is based on a study of elderly CKD white patients over 70 years of age [21]. In this study, neither preoperative nor postoperative renal function assessment has good applicability.

Conclusion
For renal function assessment in patients with renal tumors before surgery, eGFR1-14 are all significantly correlated. Among them, the MDRD-I formula can more accurately estimate the GFR of patients. In similar formulas, there is no accurate conclusion about which is better for creatinine or cystatin. In the CKD-EPI formula, the formula of using cystatin C alone is more consistent. While in the FAS formula, the combination of the two is required to calculate GFR more accurately. After surgery, the MDRD-related formula and Schwartz Scr-SCysC are not applicable in the calculation of glomerular filtration rate. Among them, CKD-EPI Scr-SCysC has good consistency in the assessment of postoperative GFR in patients with renal cancer. e FAS Scr-SCysC formula has the highest accuracy and precision. Both of these are recommended for clinical use. However, in this study, due to the small number of postoperative patients, large-scale and sample size clinical studies can still be carried out in the future to confirm it. Data Availability e data used to support the findings of this study are available from the corresponding author upon request.

Ethical Approval
e study was approved from by the Ethics Committee of Fujian Provincial Hospital.
e study was performed in accordance with the ethical standards as laid down in the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards.

Consent
Informed consent was obtained from all subjects.

Conflicts of Interest
e authors declare that they have no conflicts of interest.

Authors' Contributions
Qiuyan Li collected and analyzed data and wrote and edited the manuscript. Zesong Yang collected and analyzed data and edited the manuscript. Shiwen Zheng statistically analyzed data and wrote the manuscript. Wanghai Cai, Minxiong Hu, Yangbiao Wu, and Qingguo Zhu collected data. Liefu Ye designed and implemented the research.