Subject data
We retrospectively analyzed data of 132 non-metastatic UTUC patients between 2002 and 2010, and 61 non-metastatic RCC patients between 2003 and 2011. All patients had undergone unilateral nephrectomy through either open or laparoscopic approach at our hospitals. Parameters of age, sex, smoking, Chinese herb use, and prevalence of hypertension, diabetes mellitus, hyperlipidemia, hydronephrosis, and kidney stones were recorded. We excluded subjects who had incomplete clinical information, received renal replacement therapy preoperatively, had no pathological evidence of UTUC, and had undergone surgery twice for UTUC. Location of tumor defined as either ureter or renal pelvis based on dominant tumor features, in a sequential order of the stage, grade, and size. The renal histopathological parameters were investigated by 3 specialists: 2 nephrologists and a pathologist. Moreover, the subjects were stratified into quartiles according to age (≤54, 55–64, 65–74, and ≥ 75 years) and sex for pre-existing CKD prevalence and other analyses. The flow chart is described in Fig. 1. The study protocol was approved by our Institutional Review Board (KMUH-IRB-20120138).
Pre-existing CKD evaluation
To evaluate the patient’s pre-operation kidney function, the latest creatinine level obtained within 30 days preoperatively was collected. We used CKD Epidemiology Collaboration Equation (CKD-EPI) to calculate the estimated glomerular filtration rate (eGFR) [11].
$$ \mathrm{eGFR}=141\times \min\ {\left(\mathrm{Scr}/\upkappa, 1\right)}^{\upalpha}\times \max\ {\left(\mathrm{Scr}/\upkappa, 1\right)}^{\hbox{-} 1.209}\times 0{.993}^{\mathrm{Age}}\times 1.018\left(\mathrm{if}\ \mathrm{female}\right)\times 1.159\left(\mathrm{if}\mathrm{black}\right), $$
Were Scr is the serum creatinine, α is − 0.329 for women and − 0.411 for men, κ is 0.7 for women and 0.9 for men, min indicates the minimum Scr/κ or 1, and max indicates the maximum Scr/κ or 1. The pre-existing CKD stages of all patients were determined based on their pre-existing eGFR at the time of unilateral nephrectomy. All the patients were stratified into stages of CKD based on the Kidney Dialysis Outcomes Quality Initiative (K-DOKI) classification, as follows: stage 1, GFR > 90 mL/min/m2 with proteinuria or microalbuminuria; stage 2, GFR 60–89 mL/min/m2 with proteinuria or microalbuminuria; stage 3A, GFR = 45–59 mL/min/m2; stage 3B, GFR = 30–44 mL/min/m2; stage 4, GFR = 15–29 mL/min/m2; and stage 5, GFR < 15 mL/min/m2.
Histochemical staining
Kidney specimen were thoroughly excision and representation, section are taken from the non-tumorous area, which is at least 1 cm distance from the tumor. Formalin-fixed paraffin embedded (FFPE) blocks of the non-tumorous kidney parenchyma were retrieved. Tissue sections of 3 μm were cut, deparaffinized, and rehydrated. Hematoxylin-eosin (H&E) stain, periodic acid-Schiff (PAS) stain and Masson trichrome stain were performed as recommended [12].
Pathological evaluation
The global glomerulosclerosis (GGS) rate and tubulointerstitial (TI) score were semiquantified by 2 nephrologists and one pathologist, who were blinded to the patients’ clinical information. For cases with discrepancy, a consensus was made after review the slides together at a multi-headed microscope. The TI score was the sum of the severity level of four pathological features: tubular necrosis (Fig. 2a; 0: normal tubules, 1: rare single necrotic tubule, 2: several clusters of necrotic tubules, and 3: confluence of necrotic clusters), tubular atrophy (Fig. 2b; 0: normal tubules, 1: rare single atrophic tubule, 2: several clusters of atrophic tubules, and 3: confluence of atrophic tubular clusters), lymphocytic infiltrates (Fig. 2c; 0: absent, 1: few scattered cells, 2: group of lymphocytes, 3: and widespread infiltrates), and interstitial fibrosis (Fig. 2d; 0: absent, 1: minimal fibrosis, 2: moderate fibrosis, and 3: severe fibrosis), ranging from 0 to 12 [13]. GGS rate was the number of glomeruli with global glomerulosclerosis, defined as glomerulus with more than 50% of area involved by sclerosis, over the number of the glomeruli that can be found in the slides (Fig. 2e). Since GGS developed as an individual getting old, we compared the observed GGS rate with the estimated GGS, calculated by using an equation, (age X 0.5) – 10, that proposed by Smith et al. [14]. If the observed GGS rate exceeded the estimated GGS, it was considered as “abnormal GGS rate” (Table 3). For an example, a 40% of observed GGS in a 80 years old patient (estimated GGS is 80 × 0.5–10 = 30%) was considered abnormal.
Postoperative follow-up
All the patients were followed by performing cystoscopic examination every 3 months in the first 2 years after nephrectomy, every 6 months in the next 2 years, and annually thereafter. During surveillance, physical examinations and cystoscopic, urine cytological, and periodic imaging studies were performed following institutional guidelines. Intraluminal recurrence was defined as the recurrence of tumors in the contralateral upper urinary tract or bladder. Metastatic progression was defined as tumor recurrence in the tumor bed or regional lymph nodes and distant metastasis.
End points
The primary end point was renal outcomes, defined as creatinine doubling or dialysis. The secondary end point was all-cause mortality. If the patients died within 3 months of the primary end point, they were not defined as having the primary end point.
Statistical analysis
Data were described as mean ± standard deviation (SD), frequency, or percentage. Student’s t test or one-way analysis of variance (ANOVA) was used for comparing the continuous variables between different groups and chi-squared test was used for comparing different distribution of categorical data. Multiple binary logistic regression was applied to explore factors associated with preexisting CKD-EPI and abnormal of GGS rate. Factors associated with TI score were evaluated by multiple linear regressions. We calculated follow-up time as time between the date of unilateral nephrectomy and the date of dialysis or creatinine doubling. The Kaplan-Meier method was used to estimate renal survival rates of the histological GGS normal and abnormal groups and test the difference between these two groups by log-rank test.
Because our patients were more likely to die than to reach renal outcomes, a competitive risk Fin-Gray regression model was used to identify independent associated predictors. All independent variables were included univariable analysis and selected into multivariable analysis under criteria of p < 0.1. All statistical analyses were performed using SPSS Version 19 (IBM, Armonk, NY, USA) or SAS 9.4 (SAS Institute Inc., Cary, NC,USA), and figures were made using GraphPad Prism 5.0 (GraphPad Software, Inc., California, USA). In all analyses, two-sided p < 0.05 was considered statistically significant.
