Paraneoplastic and symptomatic score for prediction of overall and cancer-specific survival in patients with renal cell carcinoma.

Abstract

527 Background: Patients presenting with symptomatic renal cell carcinoma (RCC) tend to have more aggressive disease and poorer prognosis. Paraneoplastic syndromes (PNS), manifesting as atypical symptoms or abnormal laboratory values, are similarly associated with worse outcomes. However, the impact of symptoms and PNS on prognosis remains unclear. The aim of this study is to develop a Paraneoplastic and Symptomatic Score (PNSS) to better predict prognosis by considering clinical presentation and the presence of PNS at diagnosis. Methods: Prospectively collected data from 5743 T1-T4, N0/1, M0/1 RCC patients treated with surgery across four centers (UC San Diego, IRCCS San Raffaele, Emory University, Tokyo Medical and Dental University) were retrospectively analyzed. The PNSS was developed using preoperative variables: hematuria, flank pain, fever, nausea/vomiting/appetite loss, anemia (<11.5 g/dL for women and <12.5 g/dL for men), hypoalbuminemia (<3.5 g/dL), hypercalcemia (>10.5 mg/dL), elevated De Ritis ratio (AST/ALT>1.25), elevated C-reactive protein (>5 mg/dL), polycythemia (hematocrit >47% for women, >53% for men) and leukocytosis (WBC>11x10^9/L). Patients were stratified into four categories: Low PNSS=0; Favorable-intermediate=1-2; Unfavorable-intermediate=1-2 + Anemia; High PNSS ≥3. Multivariable Cox regression analyses (MVA) tested the association of PNSS categories with overall (OS) and cancer-specific survival (CSS). Kaplan-Meier analysis (KMA) was used to show survival differences among groups. Receiver operating characteristic (ROC) curves and area under the curve (AUC) were used to assess the score's predictive validity for OS and CSS. Results: Of the 5743 patients, 48% (n=2770) were classified as Low PNSS, 29% (n=1656) as Favorable-intermediate, 17% (n=982) as Unfavorable-intermediate and 6% (n=335) as High PNSS. At MVA, increasing PNSS was significantly associated with worse OS and CSS (all p<0.05). High PNSS patients had the highest hazard ratios (HR) for OS [HR=6.5; 95% Confidence Interval (95CI)=4.4-9.5; p<0.05) and CSS (HR=9.6; 95CI=5.4-16.8; p<0.05). KMA showed a clear separation of survival curves across PNSS categories (log-rank p<0.001). Five-year OS rates were 92% (95CI=91-94) for Low, 84% (82-86) for Favorable-intermediate, 68% (65-72) for Unfavorable-intermediate and 47% (42-54) for High, while 5-year CSS rates were 96% (95-97), 91% (89-93), 79% (76-82) and 60% (53-67), respectively. ROC AUCs were 0.70 for OS and 0.71 for CSS, indicating strong predictive accuracy. Conclusions: Increasing PNSS scores were significantly associated with worse oncological outcomes. The PNSS represents a comprehensive tool for predicting overall and cancer-specific survival in RCC patients. Including PNSS in preoperative assessments aids in identifying patients at higher risk and guiding clinical decision-making.