上泌尿道泌尿上皮癌病患接受單側腎臟輸尿管切除術後腎功能之預測模型
蘇楷森1,、黃逸修1,2,、顧明軒1,2,、陳威任1,2,、黃子豪1,2,、魏子鈞1,2,、黃奕燊1,2,、
范玉華1,2,、林志杰1,2,、林子平1,2,、鍾孝仁1,2,、郭俊逸1,2,、張延驊1,2,、林登龍1,2,、黃志賢1,2,
1臺北榮總泌尿部；2國立陽明交通大學醫學系泌尿學科
Establishing Predictive Model of Renal Function after Unilateral Nephroureterectomy in Upper Tract Urothelial Carcinoma Patients
Kai-Sen Su1, Eric Yi-Hsiu Huang1,2,, Ming-Hsuan Ku1,2,, Wei-Ren Chen1,2,, Tzu-Hao Huang1,2,, Tzu-Chun Wei1,2,, I-shen Huang1,2,, Yu-Hua Fan1,2,, Chih-Chieh Lin1,2,, Tzu-Ping Lin1,2,,
Hsiao-Jen Chung1,2,, Junne-Yih Kuo1,2,, Yen-Hwa Chang1,2,, Alex T.L. Lin1,2,,
William J.S. Huang1,2,
1 Department of Urology, Taipei Veterans General Hospital;
2 Department of Urology, School of Medicine and Shu-Tien Urological Science Research Center, National Yang Ming Chiao Tung University, Taipei, Taiwan
Purpose: Radical nephroureterectomy with bladder cuff excision (RNU BCE) is the standard of treatment for non-metastatic upper tract urothelial carcinoma (UTUC). Aside from oncological outcomes, the impact of surgery on post-operative residual renal function is also a major concern. We hereby present a predictive model for post-unilateral RNU BCE renal function, developed with pre-operative renal scintigraphy.
Materials and Methods: From January 2010 to December 2020, patients who were diagnosed of non-metastatic UTUC and underwent unilateral RNU BCE in our institute were reviewed. Demographic data was collected. Comprehensive renal function evaluation by technetium-99m mercaptoacetyltriglycine (Tc-99m MAG3) renal scintigraphy was done pre-operatively. Serum creatinine level was recorded pre-operatively and 3 months post-operatively. Subgroup analysis for those who needed dialysis post-operatively was conducted with univariate and multivariate analysis. Pearson’s linear correlation analysis was used for developing predictive model.
Results: In total, 487 patients who underwent unilateral RNU BCE for non-metastatic UTUC were included. The median age at surgery was 73.0 years (IQR 66.0-80.0). The mean pre-operative creatinine level was 1.25 mg/dl, and the estimated glomerular filtration rate (eGFR) was 50.46 ml/min/1.732m2. The mean creatinine level was 1.74 mg/dl, and the estimated glomerular filtration rate (eGFR) was 37.94 ml/min/1.732m2 at 3 months post-operatively. The mean pre-operative effective renal plasma flow (ERPF) was 252.91 ml/min in total, and the lesion side accounted for 33.2% of total ERPF in average. Thirty patients needed dialysis post-operatively, the mean latency time of dialysis after surgery is 3.1 years (IQR 0.6-4.4). Pre-operative hemoglobin level, operation method, blood transfusion, and pre-operative lesion side ERPF ratio served as independent risk factors for post-operative dialysis. Pre-operative lesion side ERPF ratio was significantly associated to post-operative decreased eGFR ratio (R2 = 0.376, p < 0.001), and the predictive model was established as: decreased EGFR% = 0.675 x lesion site ERPF% + 0.919. Even more significant association was observed during subgroup analysis. For subgroup of pre-operative total ERPF < 150 ml/min, decreased eGFR% = 0.794 x lesion site ERPF% - 2.945 (R2 = 0.444 P < 0.001).
Conclusion: Risk factors for post-operative dialysis were identified from our database. We established predictive model using pre-operative lesion side ERPF ratio to predict post-operative decreased eGFR ratio. Stronger predictive models were established for subgroup of pre-operative total ERPF < 150 ml/min. Such equation can aid in the clinical decision and pre-operative planning for patients undergoing unilateral nephroureterectomy.