Chronic Kidney Disease Is More Common in Locally Advanced Renal Cell Carcinoma

Accepted Manuscript Title: Chronic Kidney Disease is More Common in Locally-Advanced Renal Cell Carcinoma Author: Sumi Dey, Zachary Hamilton, Sabrina L. Noyes, Conrad M. Tobert, Jacob Keeley, Ithaar H. Derweesh, Brian R. Lane PII: DOI: Reference:

S0090-4295(17)30296-0 http://dx.doi.org/doi: 10.1016/j.urology.2017.03.033 URL 20367

To appear in:

Urology

Received date: Accepted date:

20-2-2017 23-3-2017

Please cite this article as: Sumi Dey, Zachary Hamilton, Sabrina L. Noyes, Conrad M. Tobert, Jacob Keeley, Ithaar H. Derweesh, Brian R. Lane, Chronic Kidney Disease is More Common in Locally-Advanced Renal Cell Carcinoma, Urology (2017), http://dx.doi.org/doi: 10.1016/j.urology.2017.03.033. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Chronic kidney disease is more common in locally-advanced renal cell carcinoma Sumi Dey1, Zachary Hamilton2, Sabrina L. Noyes1, Conrad M. Tobert3, Jacob Keeley1, Ithaar H. Derweesh2, Brian R. Lane1,4

1

Spectrum Health, Grand Rapids, Michigan

2

University of California San Diego, San Diego, California 3

University of Iowa, Iowa City, Iowa

4

Michigan State University College of Human Medicine, Grand Rapids, Michigan

*Reprints and correspondence: Brian R. Lane, M.D., Ph.D., FACS Betz Family Endowed Chair for Cancer Research, Spectrum Health Chief, Urology, Spectrum Health Medical Group Clinical Associate Professor, Michigan State University College of Human Medicine 4069 Lake Drive, Suite 313, MC: 9016 Grand Rapids, MI 49546 Tel: 616.267.7333 Fax: 616.267.8040 E-mail: [email protected] Abstract word count: 235 Word count: 2137

Tables: 3 Figures: 1 References: 29

Running title: Association between CKD and Renal Cell Carcinoma

Conflicts of Interest: The authors have no conflicts of interest. Keywords: Kidney cancer; chronic kidney disease, glomerular filtration rate, proteinuria Acknowledgements: The corresponding author would like to thank the Betz Family Endowment for Cancer Research for their continued support. The authors would also like to thank Jessica Parker for assistance with statistics. Funding: Funding was provided in part by the Spectrum Health Foundation (RG0813-1036). 1 Page 1 of 20

ABSTRACT

Objective: To retrospectively evaluate clinical predictors of CKD in RCC patients to identify associations between patient- and tumor-specific factors with poorer renal function. Chronic kidney disease (CKD) and renal cell carcinoma (RCC) are inter-related, with 26-44% of RCC patients having concomitant CKD at diagnosis. Methods: Institutional registries from SH and UCSD were queried for preoperative GFR and proteinuria status prior to radical or partial nephrectomy. Preoperative clinical and tumor factors were recorded; proteinuria was classified as A1 (<30 mg), A2 (30-300mg), and A3 (>300 mg). CKD was grouped by KDIGO classification (low, moderately-increased, high, very-high). Results: We evaluated 1,569 patients undergoing surgery for renal cortical tumors. CKD status was lowrisk in 860 (55%), moderately-increased in 381 (24%), high in 194 (12%), and very high in 134 (9%) patients. Increased RENAL score, tumor size, and clinical tumor stage were strongly associated with increased CKD risk at baseline. Clinical stage T3/T4 disease had more at risk patients then stage T2 and T1 disease (39.5% vs. 22% and 19%, p=0.0001). Clinical tumor stage and gender were the only predictors of proteinuria, lower GFR, and higher CKD risk group in both univariate and multivariate analysis. Conclusions: Forty-five percent of patients with RCC had moderate or higher CKD before treatment. A positive correlation between pre-treatment CKD and locally-advanced RCC (cT3/T4) was present. This likely relates to increased loss of functional parenchyma with increasing tumor size/stage with important implications in patient management.

2 Page 2 of 20

INTRODUCTION

According to the KDIGO international group recommendation, chronic kidney disease (CKD) is defined as either evidence of kidney damage (such as albuminuria, abnormalities detected by imaging, histology due to tubular disorder, or abnormal urinary sediment) or glomerular filtration rate (GFR) <60 mL/min per 1.73 m2 present for >3 months.1 Although GFR is widely accepted as the best overall index of kidney function, it is now recommended that CKD be classified based on cause, GFR, and albuminuria because of improved classification and estimation of kidney disease and progression risk.1 CKD can be due to diabetes mellitus (DM), hypertension (HTN), glomerular disease, autoimmune disease, interstitial disease, loss of functioning renal parenchyma (by replacement, such as with renal cancer or removal by surgery), and albuminuria is a sensitive marker for CKD irrespective of cause.1 The global incidence of renal cell carcinoma (RCC) is increasing annually, accounting for 2% to 3% of all adult malignant neoplasms2 with 62,700 new cases and 14,240 deaths estimated in 2016 in the United States.3 Preoperative CKD is very common in RCC patients and it has been estimated about 26-34% have CKD (GFR <60 ml/min/1.73 m2) before surgery.4,5 Previous studies have indicated proteinuria as an independent predictor of renal function decline and mortality.6

When considering proteinuria, the rate of patients with preoperative CKD

increases to 44-50%.7-9 CKD is a progressive condition and the rate of progression depends on the underlying cause, GFR, and albuminuria status.10 Our previous single-institution analysis demonstrated that decreased GFR and proteinuria are significant predictors of mortality and progression of CKD in kidney cancer patients.11 In this study, we retrospectively evaluated pretreatment KDIGO CKD status (based on proteinuria and GFR) from a multi-institutional cohort

3 Page 3 of 20

of patients with all stages of RCC to evaluate the relationship of patient and tumor-specific factors to severity of CKD.

PATIENTS AND METHODS

We conducted a multi-institutional retrospective study of all patients undergoing surgery for renal cortical tumors at Spectrum Health and University of California San Diego (UCSD). Inclusion criteria were age >18 years old, clinical evidence of solid or complex cystic renal mass (non-metastatic and metastatic), documentation of clinical TNM stage, and either radical nephrectomy (RN) or partial nephrectomy (PN).

Preoperative estimated GFR values were

calculated using the CKD epidemiology collaboration formula.12

GFR values were sub-

classified into 6 groups: G1 ≥90, G2 = 60-89, G3a = 45-59, G3b = 30-44, G4 = 15-29, and G5 <15 ml/min/1.73m2.1 Preoperative proteinuria was assessed by dipstick urinalysis and subclassified into 3 groups: A1 <30 mg/dl, A2 = 30-300 mg/dl, and A3 >300 mg/dl.1 According to KDIGO guidelines, CKD was categorized into four risk groups (low, moderately-increased, high, and very-high) by using GFR level (G1-G5) and albuminuria (A1-A3) (Appendix).1 Patients from Spectrum Health (n=1,087) and UCSD (n=482) meeting inclusion criteria and having both preoperative GFR and proteinuria status prior to surgery comprised the study group. Pathological tumor stage was classified according to the international TNM staging system for RCC.2 Continuous variables were reported as the median with interquartile ratio (IQR: 25th, 75th percentiles) or as the median ± standard deviation (SD). Differences between categorical variables were determined using Chi-square test.

Statistical analyses were run on the

4 Page 4 of 20

retrospective cohort to determine predictors of proteinuria group (A1-A3), GFR group (G1-G5), and KDIGO group. Univariate logistic regression model was fit to the proteinuria, GFR group, and KDIGO group using age, sex, race, HTN, DM, CAD, tumor (side, size, grade), cancer versus benign, RENAL score, and pathological stage as covariates. Multivariable models were fit to the same predictors using the indicated covariates. All statistical analyses were generated using SAS Enterprise Guide software, Version 7.1 © 2014 SAS Institute Inc. Statistical significance was assessed at p <0.05.

RESULTS

Based on the inclusion criteria, the two retrospective cohorts from SH and UCSD were combined for a total of 1,569 patients undergoing surgery for renal cortical tumors. Table 1 indicates the demographics of the patient cohort based on clinical stage. Median GFR was 68.5 ml/min/1.73m2 (IQR: 53-87).

Proteinuria was present in 20% of the patients at baseline

including 268 (17%) with 30-300 mg/dl (A2) and 45 (3%) with >300 mg/dl (A3). Categorization of patients according to KDIGO-defined CKD at baseline identified 860 patients (55%) as low risk, 381 (24%) as moderately-increased risk, 194 (12%) as high-risk, and 134 (9%) as very-high risk. Tumor features grouped according to preoperative CKD are shown in Table 2. Median tumor size was 3.7 cm (IQR: 2.4, 6.2) and median RENAL score was 8 (IQR: 6, 10). Eighty seven percent of the final pathology was malignant with 62% of the total cohort diagnosed with clear cell morphology. One thousand two hundred thirty two patients (79%) had cT1 tumors, 261 (17%) had cT2 tumors, and 76 (5%) had cT3/T4 tumors.

5 Page 5 of 20

Associations between tumor characteristics and baseline KDIGO-defined CKD are illustrated in Figure 1.

Increased clinical tumor stage (p<0.0001), tumor size (p<0.0022),

Robson clinical stage13 (p<0.0001), and Fuhrman nuclear grade (p=0.0469) were correlated with higher KDIGO-defined CKD risk. Specifically for Robson clinical stage (Figure 1D), stage III had more at risk patients then stage IV patients: 38% and 27.5% of patients, respectively, were categorized as high or very-high risk (p=0.33). Although patients with cancer were no more likely to have preexisting CKD (Figure 1F), higher tumor grade was associated with KDIGOdefined CKD (Figure 1E). When patients were evaluated according to RCC subtype (data not shown), papillary RCC had the greatest likelihood of preoperative CKD; the proportion of patients with high or very-high risk CKD was 32%, 18%, 16%, and 18% for papillary RCC, clear cell RCC, chromophobe RCC, and benign renal tumors, respectively (p=0.0008). Table 3 shows univariate and multivariate models for predictors of kidney disease as evidenced by proteinuria, reduced GFR, and KDIGO classification. RENAL score was omitted from multivariate analyses due to the large amount of missing data. On univariate analysis, Robson clinical stage and clinical tumor stage were the strongest predictors of proteinuria; chisquared 32.74 and 31.71 respectively (p<0.0001). Gender, race, tumor size, and tumor grade were also significant predictors. In a multivariate model of proteinuria, gender, race and clinical tumor stage were all statistically significant predictors. In a univariate analysis predicting CKD based on GFR, age and gender were the strongest predictors, chi-squared 320.84 and 59.03 respectively (p<0.0001). Tumor size, HTN, tumor grade, clinical tumor stage, RENAL score, and Robson clinical stage were also significant predictors. In a multivariate model of CKD based on GFR, age, gender, HTN, coronary artery disease (CAD), and clinical tumor stage were all significant predictors. In a univariate analysis predicting KDIGO risk classification, age and

6 Page 6 of 20

Robson clinical stage were the strongest predictors, chi-squared 140.70 and 31.33 respectively (p<0.0001). Gender, race, HTN, tumor size, tumor grade, clinical tumor stage, and RENAL score were also significant predictors. In a multivariate model of KDIGO risk classification, age, gender, race, tumor size, and clinical tumor stage were all significant predictors. Due to the high correlation (0.96) between Robson clinical stage and clinical tumor stage, including both of them in the multivariate model would be inappropriate and would not bring any new information to the model. With all other predictors in the model, clinical tumor stage proves to be more significant than Robson stage and was; therefore, selected to stay in the model. Only gender and clinical tumor stage were significant predictors of proteinuria, reduced GFR, and CKD by KDIGO classification in univariate and each multivariate analysis. Clinical tumor stage outperformed gender in multivariate models of proteinuria and KDIGO risk classification (Table 3). Clinical T stage was strongly associated with preoperative CKD status (p=0.0001), with only 19% of patients with stage T1 having high or very-high risk CKD. Despite having similarly sized tumors, patients with cT3/T4 were significantly more likely to have high or very-high risk CKD then those with cT2 disease (39.5% vs. 22%, p=0.0026). In addition, of the 69 patients with clinical stage IV disease, 38 (55.1%) were < cT2 tumors. The odds of having high or very-high risk CKD is 2.62 times higher for patients with cT3/T4 tumors when compared to patients with cT1/T2 tumors (39.5% vs. 20.0%, p=<0.0001).

DISCUSSION

CKD is a common comorbidity amongst patients undergoing surgery for renal cortical tumors.4,5,7-9,14

Understanding this association and early recognition is of even greater

7 Page 7 of 20

importance given the impact of preexisting CKD on overall and non-cancer related mortality and the functional deterioration associated with treatment.14-19 Despite concerns regarding iatrogenic creation or worsening of preexisting CKD, PN continues to be grossly underutilized for small renal masses.20 Traditionally, CKD has been defined using serum creatinine and more recently with Modification of Diet in Renal Disease (MDRD) and Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI), because these equations incorporate patient age, gender, body-surface area, and race.12 Canter et al previously quantified preoperative CKD in patients undergoing surgery for renal cortical tumors using the MDRD and CKD-EPI equations.21 Of 1,114 patients, 22% of patients had CKD stage III or higher. In contrast, our study found that 36% of patients presented with CKD stage III or higher. Expanding the stratification of CKD risk by using the KDIGO classification, led to the identification of only 55% of patients at low-risk and 21% at high or very-high risk prior to surgery for renal cortical tumors. Increased tumor burden leads to a decrease in functional nephrons. Our study provides a unique perspective to demonstrate this phenomenon.

In both univariate and multivariate

analysis, clinical tumor stage was the only characteristic that was a significant predictor of proteinuria, GFR, and KDIGO risk classification (Table 3). Tumor size and tumor grade were both significant predictors in univariate analysis. Interestingly, when comparing clinical tumor stage III and IV, patients with stage III tumors were more likely to have pre-treatment CKD then those with stage IV disease (Figure 1D). This is likely due to the increase in local tumor burden compromising functional nephrons. Of the patients with clinical stage IV disease, 38 (55%) were < cT2 tumors. Those patients with cT3/T4 tumors were more likely to have preoperative CKD. These findings are consistent with prior literature indicating that 38% of stage IV renal cancer patients had < cT2 tumors.22

8 Page 8 of 20

The presence of proteinuria prior to treatment for kidney cancer can stratify patients according to current and future risk of CKD progression and mortality.1,23 We recently reported that the presence of preoperative proteinuria predicted renal functional decline and was also associated with decreased overall survival in patients with renal cortical tumors.9,11 Examining tumor characteristics that were associated with increased proteinuria, clinical tumor stage, and tumor size were the strongest predictors in univariate analysis, chi-squared 31.71 and 17.13 respectively (p<0.0001).

In a multivariate analysis, clinical tumor stage remained the sole

predictor of proteinuria. Vaglio et al evaluated the prognostic significance of proteinuria in patients with renal cortical tumors before and after receiving immunotherapy.24 Similar to our findings here, before treatment, the presence of proteinuria was more common in patients with stage III/IV as compared to stage I/II, p=0.03. The etiology of decreased renal function in patients with locally advanced disease is difficult to determine. Multiple studies have shown the correlation between preservation of functional nephrons and post-operative renal function.19,25-27 In our study, we found that patients with stage III disease had worse renal function then stage II (despite similarly sized tumors). The loss of functional parenchyma with increased tumor size and complexity, along with the disruption of blood flow to the affected kidney is a likely etiology specifically in patients with tumor thrombus (stage III).28 There is significant evidence in the literature suggesting that reduced blood flow leading to intrarenal ischemia may contribute to fibrosis in renal parenchyma.29,30 Despite the known development of collateral blood flow in patients with renal vein obstruction, the long-term effects of ischemia to the remaining nephrons is difficult to quantify, and these patients may be more likely to recover renal function after nephrectomy.31 In the setting of locally advanced RCC, it is imperative for urologists to be mindful of preoperative

9 Page 9 of 20

renal function and potential post-operative outcomes in counseling of patients. As the involved kidney may have reduced function, a renal scan to assess relative renal function is likely indicated in many such patients to help make clinical decisions.

We acknowledge several limitations of our study. Our study is a retrospective analysis from two institutions and we did not have a complete array of tumor characteristics for all patients, most notably RENAL score which was not available for all patients. In order to determine KDIGO classification, proteinuria was assessed by urinary dipstick. Urinary dipstick is less sensitive and less reliable than spot albumin-to-creatinine ratio (ACR) and 24-hour assessment of urinary protein.

In addition, proteinuria detected on urinalysis may be less

accurate in the presence of hematuria or pyuria; however, guidelines regarding this issue are lacking.

CONCLUSION

Based on KDIGO classification, 45% of patients undergoing treatment for renal cortical tumors have baseline CKD. A positive correlation between pre-treatment CKD and locallyadvanced RCC (cT3/T4) was present.

This likely relates to increased loss of functional

parenchyma and hemodynamic influences of locally-advanced disease with increasing tumor size/stage and has important implications in patient management.

10 Page 10 of 20

REFERENCES

1.

Kidney Disease: Improving Global Outcomes Chronic Kidney Disease Guideline Development Work Group M: KDIGO clinical practice guideline for the evaluation and management of chronic kidney disease. Kidney Int Suppl. 2013;3:1-150.

2.

Campbell SC and Lane BR: Malignant renal tumors., in Wein AJ, Kavoussi LR, Partin AW and Peters CA: Cambell-Walsh Urology Philadelphia, PA, Saunders, 2016, vol. 2, pp 1314-1364.

3.

Siegel RL, Miller KD and Jemal A: Cancer statistics, 2016. CA Cancer J Clin. 2016;66:7-30.

4.

Huang WC, Levey AS, Serio AM, et al: Chronic kidney disease after nephrectomy in patients with renal cortical tumours: a retrospective cohort study. Lancet Oncol. 2006;7:735-740.

5.

Stevens LA, Li S, Wang C, et al: Prevalence of CKD and comorbid illness in elderly patients in the United States: results from the Kidney Early Evaluation Program (KEEP). Am J Kidney Dis. 2010;55:S23-33.

6.

Drury PL, Ting R, Zannino D, et al: Estimated glomerular filtration rate and albuminuria are independent predictors of cardiovascular events and death in type 2 diabetes mellitus: the Fenofibrate Intervention and Event Lowering in Diabetes (FIELD) study. Diabetologia. 2011;54:32-43.

7.

Hamilton Z, Dey S, Berquist S, et al: Should partial nephrectomy be considered an imperative indication in stage II chronic kidney disease?: American Urological Association. San Diego, CA, 2016.

11 Page 11 of 20

8.

Campbell SC, Dong W, Zabell J, et al: End-stage Renal Disease after Renal Surgery: Partial Nephrectomy is Protective, but to What Degree and Consequence? Eur Urol. 2016.

9.

O'Donnell K, Tourojman M, Tobert CM, et al: Proteinuria is a Predictor of Renal Functional Decline in Patients with Kidney Cancer. J Urol. 2016.

10.

Snyder S and Pendergraph B: Detection and evaluation of chronic kidney disease. Am Fam Physician. 2005;72:1723-1732.

11.

Tourojman M, Kirmiz S, Boelkins B, et al: Impact of Reduced Glomerular Filtration Rate and Proteinuria on Overall Survival of Patients with Renal Cancer. J Urol. 2016;195:588593.

12.

Levey AS, Stevens LA, Schmid CH, et al: A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150:604-612.

13.

Robson CJ, Churchill BM and Anderson W: The results of radical nephrectomy for renal cell carcinoma. J Urol. 1969;101:297-301.

14.

Lowrance WT, Ordonez J, Udaltsova N, et al: CKD and the risk of incident cancer. J Am Soc Nephrol. 2014;25:2327-2334.

15.

Hu SL, Chang A, Perazella MA, et al: The Nephrologist's Tumor: Basic Biology and Management of Renal Cell Carcinoma. J Am Soc Nephrol. 2016.

16.

Lane BR, Poggio ED, Herts BR, et al: Renal function assessment in the era of chronic kidney disease: renewed emphasis on renal function centered patient care. J Urol. 2009;182:435-443; discussion 443-434.

12 Page 12 of 20

17.

Lane BR, Demirjian S, Weight CJ, et al: Performance of the chronic kidney diseaseepidemiology study equations for estimating glomerular filtration rate before and after nephrectomy. J Urol. 2010;183:896-901.

18.

Lane BR, Campbell SC, Demirjian S, et al: Surgically induced chronic kidney disease may be associated with a lower risk of progression and mortality than medical chronic kidney disease. J Urol. 2013;189:1649-1655.

19.

Weight CJ, Larson BT, Fergany AF, et al: Nephrectomy induced chronic renal insufficiency is associated with increased risk of cardiovascular death and death from any cause in patients with localized cT1b renal masses. J Urol. 2010;183:1317-1323.

20.

Huang WC, Atoria CL, Bjurlin M, et al: Management of Small Kidney Cancers in the New Millennium: Contemporary Trends and Outcomes in a Population-Based Cohort. JAMA Surg. 2015;150:664-672.

21.

Canter D, Kutikov A, Sirohi M, et al: Prevalence of baseline chronic kidney disease in patients presenting with solid renal tumors. Urology. 2011;77:781-785.

22.

Richey SL, Culp SH, Jonasch E, et al: Outcome of patients with metastatic renal cell carcinoma treated with targeted therapy without cytoreductive nephrectomy. Ann Oncol. 2011;22:1048-1053.

23.

Wang HE, Jain G, Glassock RJ, et al: Comparison of absolute serum creatinine changes versus Kidney Disease: Improving Global Outcomes consensus definitions for characterizing stages of acute kidney injury. Nephrol Dial Transplant. 2013;28:14471454.

24.

Vaglio A, Buzio L, Cravedi P, et al: Prognostic significance of albuminuria in patients with renal cell cancer. J Urol. 2003;170:1135-1137.

13 Page 13 of 20

25.

Tobert CM, Boelkins B, Culver S, et al: Surgeon assessment of renal preservation with partial nephrectomy provides information comparable to measurement of volume preservation with 3-dimensional image analysis. J Urol. 2014;191:1218-1224.

26.

Lane BR, Fergany AF, Weight CJ, et al: Renal functional outcomes after partial nephrectomy with extended ischemic intervals are better than after radical nephrectomy. J Urol. 2010;184:1286-1290.

27.

Sun M, Bianchi M, Hansen J, et al: Chronic kidney disease after nephrectomy in patients with small renal masses: a retrospective observational analysis. Eur Urol. 2012;62:696703.

28.

Witz M, Kantarovsky A, Morag B, et al: Renal vein occlusion: a review. J Urol. 1996;155:1173-1179.

29.

Palm F and Nordquist L: Renal oxidative stress, oxygenation, and hypertension. Am J Physiol Regul Integr Comp Physiol. 2011;301:R1229-1241.

30.

Hill GS: Hypertensive nephrosclerosis. Curr Opin Nephrol Hypertens. 2008;17:266-270.

31.

Zabor EC, Furberg H, Mashni J, et al: Factors Associated with Recovery of Renal Function following Radical Nephrectomy for Kidney Neoplasms. Clin J Am Soc Nephrol. 2016;11:101-107.

14 Page 14 of 20

Figure 1. Preoperative KDIGO-defined CKD according to: (A) clinical tumor stage, (B) tumor size, (C) tumor complexity, (D) Robson clinical stage, (E) Fuhrman nuclear grade, and (F) benign or malignant histology. Tumor size was grouped (≤4.0 cm, 4.1-7.0 cm, 7.1-10 cm, >10 cm) and was grouped according to RENAL nephrometry score (low 4-6; medium 7-9; high 10-12). Data were available for tumor stage (n=1,569), tumor size (n=1,550), complexity (n=483), clinical stage (n=1,569), grade (n=1,224), and histology (n=1,550).

15 Page 15 of 20

Table 1. Demographic data regarding patients grouped according to clinical stage. Total Patients 1569 62 (52,71) 5.8% (91) 58% (907) 3.7 (2.4, 6.2)

cT1 79% (1232) 62 (52,72) 84% (76) 77% (694) 3.1 (2.0,4.5)

cT2 17% (261) 61.5 (53,70) 16% (15) 18% (166) 9.0 (7.9,10.6)

Total Patients Median Age, years (IQR) African American Male Median Tumor Size, cm (IQR) Comorbidities HTN 53% (818) 53% (639) 54% (141) DM 43% (666) 43% (519) 46% (119) CAD 35% (531) 35% (417) 37% (95) Renal Function Median GFR (IQR) 68.5 (52.6,86.8) 70.3 (53.5,87.8) 64.9 (52.5,83.3) GFR Group G1 21% (323) 21% (264) 19% (50) G2 43% (674) 44% (540) 43% (113) G3a 21% (329) 19% (237) 23% (61) G3b 10% (158) 10% (123) 8% (22) G4 3% (54) 3% (43) 4% (9) G5 2% (31) 2% (25) 2% (6) Proteinuria Group A1 80% (1256) 82% (1014) 77% (200) A2 17% (268) 15% (189) 19% (50) A3 3% (45) 2% (29) 4% (11) KDIGO Group Low 55% (860) 58% (709) 51% (134) Moderately-increased 24% (381) 23% (283) 26% (69) High 12% (194) 11% (140) 14% (36) Very High 9% (134) 8% (100) 8% (22) Patient data were missing for hypertension (n=34), DM (n=34), and CAD (n=35).

cT3/T4 5% (76) 65 (58,72) 0% (0) 5% (47) 9.0 (7.0,10.8) 51% (38) 37% (28) 25% (19) 54.6 (44.9,67.8) 12% (9) 28% (21) 41% (31) 17% (13) 3% (2) 0% (0) 55% (42) 38% (29) 7% (5) 22% (17) 38% (29) 24% (18) 16% (12)

16 Page 16 of 20

Table 2. Tumor features grouped according to preoperative CKD defined by KDIGO.

Total Patients Tumor side (Right) Median Tumor size, cm (IQR) RENAL score (IQR)* Pathologic diagnosis* Cancer Benign Histology

Total Pts

Low Risk

ModeratelyIncreased Risk

1569

55% (860)

24% (381)

12% (194)

51% (662)

50% (353)

55% (172)

47% (78)

53% (59)

3.7 (2.4,6.2)

3.5 (2.2,5.5)

4.0 (2.4,6.9)

4.0 (2.8,7)

4.4 (2.5,7)

8 (6,10)

8 (6,10)

9 (7,10)

8 (6,10)

9 (7,10)

87% (737)

87% (325)

14% (115)

13% (50)

87% (167) 13%(24)

90% (119) 10% (13)

87% (1348) 13% (202)

High Risk

VeryHigh Risk 9% (134)

Clear cell RCC

62% (970)

64%(554)

62% (238)

Papillary RCC Chromophobe RCC Other cancer Benign Unknown Fuhrman nuclear grade* 1 2 3 4 Tumor stage

13% (211)

11% (96)

13% (48)

54% (104) 20% (38)

6% (89)

6% (50)

7% (25)

4% (7)

5% (7)

5% (74) 13% (202) 1% (23)

4% (36) 13% (115) 1% (9)

4% (14) 13% (50) 2% (6)

8% (16) 12% (24) 3% (5)

6% (8) 10% (13) 2% (3)

10% (123) 48% (584) 34% (410) 9% (107)

11% (75) 51% (350) 31% (212) 7% (45)

8% (24) 43% (125) 37% (110) 12% (35)

10% (14) 44% (64) 35% (50) 11% (16)

10% (10) 43% (45) 37% (38) 11% (11)

82% (709)

74% (283)

16% (134) 2% (17)

18% (69) 8% (29)

72% (140) 19% (36) 9% (18)

75% (100) 16% (22) 9% (12)

82% (705)

73% (277)

14% (121) 1% (11) 3% (23)

14% (55) 6% (22) 7% (27)

cT1 cT2 cT3/T4 Clinical stage14 I II III IV

79% (1232) 17% (261) 5% (76) 78% (1218) 15% (229) 3% (53) 4% (69)

72% (139) 16% (32) 7% (13) 5% (10)

55% (74) 22% (29)

72% (97) 16% (21) 5% (7) 7% (9) 17 Page 17 of 20

Data was missing for pathologic diagnosis (n=19), RENAL score (n=1086), Fuhrman nuclear grade (n=345).

18 Page 18 of 20

Table 3. Predictors of preoperative proteinuria and CKD using logistic regression. Predictors of Proteinuria group (A1A3) Univariate Multivariate (n=1569*) (n=1160) ChiP-Value ChiPSquared Squared Value 0.39 0.53 (n=1565) Gender 8.69 0.0032 (Male) (n=1569) African15.17 <0.0001 American (n=1569) 0.31 0.58 Hypertension (n=1535) Diabetes 2.37 0.12 mellitus (n=1535) Coronary 0.004 0.95 artery (n=1534) disease 0.29 0.59 Tumor side (n=1301) 17.13 <.0001 Tumor size (n=1484) 11.0 0.012 Tumor grade (n=1224) Clinical 31.71 <.0001 tumor stage (n=1569) (OR) Cancer (vs. 2.91 0.088 Benign) (n=1550) Age

1.30

0.254

4.70

0.030

17.56

<.0001

1.57

0.210

2.76

0.096

1.72

0.190

--

--

2.34

0.126

0.472

0.924

8.39

0.015

--

--

Predictors of GFR group (G1-G5) Univariate Multivariate (n=1569*) (n=1160) ChiP-Value ChiP-Value Squared Squared

Predictors of KDIGO group (LowVery High) Univariate Multivariate (n=1569*) (n=1160) ChiP-Value Chi PSquared Squared Value

320.84 (n=1565) 59.03 (n=1569) 0.0984 (n=1569) 26.52 (n=1535) 2.48 (n=1535) 0.0213 (n=1534)

<.0001

223.59

<.0001

<.0001

37.00

<.0001

0.7542

3.50

0.0613

<.0001

7.13

0.0076

0.1151

0.097

0.7555

0.8841

7.2538

0.0071

140.70 <0.0001 (n=1565) 25.17 <0.0001 (n=1569) 6.86 0.0088 (n=1569) 14.99 0.0001 (n=1535) 2.54 0.1111 (n=1535) 0.4962 0.4812 (n=1534)

0.3803 (n=1301) 4.12 (n=1484) 8.64 (n=1224) 11.71 (n=1569)

0.5374

--

--

0.042

1.0325

0.3096

0.0345

0.9996

0.8013

0.0029

8.670

0.0131

0.8624

--

--

0.030 (n=1550)

0.394 (n=1301) 16.15 (n=1484) 12.90 (n=1224) 31.02 (n=1569) 0.701 (n=1550)

95.95

<.0001

10.94

0.0009

18.54

<.0001

1.40

0.2361

0.350

0.5541

0.929

0.3351

0.5305

--

--

<.0001

5.3707

0.0205

0.0048

0.2274

0.9731

<.0001

13.88

0.0010

0.4025

--

--

19 Page 19 of 20

RENAL 1.00 0.32 --9.05 0.0026 --6.83 0.0090 -score (n=483) (n=483) (n=483) Robson 32.74 <.0001 --15.86 0.0012 --31.33 <.0001 -clinical stage (n=1569) (n=1569) (n=1569) Analyses are predictive of higher grade of proteinuria (A1, A2, A3), GFR group (G1, G2, G3a, G3b, G4, G5), and KDIGO class (low, intermediate, high, and very-high risk). Key: CKD, chronic kidney disease; GFR, glomerular filtration rate. *Number of included patients unless n is mentioned in table specifically

---

20 Page 20 of 20