Preoperative Nomogram prediction model for lymph node metastasis in patients with renal cell carcinoma / 中华泌尿外科杂志

ABSTRACT

Objective:To identify preoperative clinical predictors of positive lymph nodes in patients with renal cell carcinoma (RCC)and provide a preoperative predictive model.Methods:The data of 173 RCC patients who underwent either retroperitoneal lymph node dissection or biopsy at a single institution from January 2016 to December 2020 were retrospectively analyzed. There were 109 males and 64 females, with an average age of (53.29±13.58) years, median tumor diameter of 70 (23-150) mm, 68 patients with local symptoms, 24 patients with systemic symptoms, and 56 patients with ECOG score ≥1. There were 96 patients with tumor pseudocapsule, 23 patients with renal vein or inferior vena cava tumor thrombus, 114 patients in stage T 1-2, 59 patients in stage T 3-4, 22 patients with distant metastasis and 88 patients with lymph node metastasis by preoperative imaging examination. Univariate analysis was performed by Mann-Whitney U test or Chi-square test, and multivariate logistic regression analysis was used to determine preoperative predictors of pathologic lymph node positivity. The significant variables were then included in a novel Nomogram to predict the probability of lymph node invasion.C-index and Bootstrap self-sampling methods were used to evaluate the discrimination and consistency of the model. Results:Of the 173 patients, 49(28.32%)and 124(71.68%)had pN 1 and pN 0 disease, respectively. Among 88 patients with suspected lymph node involvement (cN 1) assessed by preoperative imaging, only 47.73%(42/88) were confirmed to be pathologically positive. However, 8.24% (7/85) of the 85 patients with negative lymph nodes (cN 0) assessed by preoperative imaging were pathologically positive. Age, ECOG score, symptoms at presentation, tumor pseudocapsule, metastasis at diagnosis, clinical tumor stage, clinical nodal status, clinical nodal size, D-dimer, lactate dehydrogenase, microscopic hematuria were significant in the univariate analysis ( P<0.05). On multivariable analyses, the most informative independent predictors were age, clinical tumor stage, clinical nodal status, clinical nodal size and microscopic hematuria ( P<0.05). A Nomogram with good performance was developed to predict the probability of lymph node metastasis. The C-index of the model was 0.954, the calibration curve of forecasting curve with the ideal curve fit was good, indicating that the model has a good consistency. Conclusions:Younger age, microscopic hematuria, suspected lymph node involvement in imaging, larger lymph node diameter and higher T stage were independent risk factors for renal cell carcinoma with lymph node metastasis. The Nomogram based on the above factors has good identification and calibration ability, which can help predict the probability of lymph node metastasis of renal cell carcinoma before surgery.