Development and validation of a novel scoring index (C-reactive protein, age, race, and tumor size) to predict renal functional decline post partial nephrectomy.
597 Background: Functional decline is a sequelae of extirpative renal surgery with potential for significant morbidity. We utilized pre-operative patient demographics, C-reactive protein, and tumor size to design and validate a novel scoring index to predict functional decline post partial nephrectomy. Methods: A multi-institutional dataset was utilized for analysis of patients with pre-operative estimated glomerular filtration rate (eGFR) > 60mL/min/1.73m2 by CKD-EPI equation. Multivariable analysis (MVA) was carried out for potential variables associated with development of post-operative chronic kidney disease (CKD) stage IIIB at last follow-up (eGFR < 45 mL/min/1.73m2). Significant variables were included in the predictive model and assigned an index score based on odds ratio. Receiver-operating-characteristic (ROC) analysis was employed to evaluate predictive validity, and bootstrapping technique was utilized to validate the model. Results: 924 patients were analyzed. 826 patients had post-operative eGFR > 45, while 111 patients had eGFR. Factors on MVA independently associated with increased risk of development of eGFR < 45 included age 65+ (OR = 2.6, p < 0.001), African-American race (OR = 2.3, p = 0.006), C-reactive protein level > 0.5mg/dL (OR = 5.3, p < 0.001), and tumor size > 4 cm (OR = 1.458, p = 0.189). For CART (C-reactive protein, Age, Race, Tumor size) score, the following values were assigned: age ( < 65 = 1, age > 65 = 3), race (non-African-American = 1, African-American = 2), tumor size ( < 4 = 1, > 4cm = 2), and CRP ( < 0.5mg/dL = 1, > 0.5mg/dL = 4). Analysis demonstrated 2.6% (12/469) of patients with a low (4-6) score had de novo eGFR < 45 postoperatively, while 35% (41/117) of patients with a high (10-11) score had de novo eGFR < 45. ROC analysis revealed AUC of 0.778, and ROC bootstrapping validation of 95 randomly selected patients revealed an AUC of 0.808. Conclusions: CART score represents a novel composite score that significantly predicts development of eGFR < 45 after surgery. This scoring system may assist in patient counseling and clinical decision making, as well as an impetus to improve outcomes in at-risk patient subgroups.