Malignancies of the kidney and renal pelvis account for 3% to 5% of all solid adult cancers . Despite the widespread use of abdominal cross-sectional imaging, which has led to a stage buy LY 2109761 with earlier diagnosis of renal tumors, 25% to 30% of patients with renal cell carcinoma (RCC) present with metastatic disease [2–4], and 10% of patients have tumor invasion into the inferior vena cava (IVC) at diagnosis [5–8]. Less than 1% of patients have tumor thrombus extension above the level of the hepatic veins (level III and level IV) [9,10]. Surgical extirpation remains the cornerstone of therapy for patients with RCC who have advanced IVC tumor thrombus extension [11–14]. Even with advanced multidisciplinary management, major complications occur in one-third of patients [15,16]. Surgical management is associated with a 5-year survival rate of 50% in patients with nonmetastatic disease in contemporary series [17–19].
The strongest predictors of long-term survival in patients with high-level tumor thrombi include the presence of distant metastases, regional nodal metastases, and tumor pathologic characteristics including grade and necrosis [17–23]. Patients with level IV thrombus have a greater incidence of postoperative complications when compared with those with level III thrombus; however, tumor thrombus height (level III vs. level IV) has not been shown to be an independent predictor of long-term oncologic outcomes [15,17]. Although pathologic features are the strongest predictors of long-term outcomes, these factors cannot be included in preoperative prognostic models owing to the infrequency of preoperative tissue diagnosis by biopsy in patients with RCC planned for nephrectomy and caval thrombectomy. Martinez-Salamanca et al.  reported a postoperative nomogram including pathologic variables for patients with all level thrombi (levels I–IV). Similarly, Nakayama et al.  reported on a preoperative prognostic model for patients with RCC who have venous thrombus; however, the study was small and included patients with less-advanced thrombi (levels 0–II). There are currently no preoperative clinical tools to guide patient selection and counseling for surgical therapy specifically in the setting of RCC with suprahepatic tumor thrombus (levels III–IV). Thus, our aim was to develop multivariable models and prognostic nomograms for prediction of survival and major postsurgical complications based on readily available preoperative variables in patients with RCC who have advanced (levels III–IV) tumor thrombus. In addition to providing individualized risk assessment, these models could have utility in the design of future clinical trials of targeted therapies in this setting.
Materials and methods
Institutional review board approval was obtained at all centers. We retrospectively identified all patients (n = 166) treated with nephrectomy and caval thrombectomy for RCC with level III or level IV IVC tumor thrombus between January 1, 2000, and June 30, 2013, at 4 tertiary centers. A total of 34 patients with incomplete data for analysis were excluded. The centers involved were Mayo Clinic Rochester (44 patients), the University of Texas MD Anderson Cancer Center in Houston (44 patients), the University of Texas Southwestern Medical Center at Dallas (25 patients), and the University of Wisconsin Hospital (19 patients). Tumor thrombus level was determined from transesophageal echocardiography or preoperative magnetic resonance imaging. Level III thrombus was defined as thrombus above the hepatic veins but below the diaphragm, and level IV thrombus was defined as thrombus extending above the diaphragm [9,10]. Management including the use of preoperative targeted therapy, preoperative angioembolization, lymphadenectomy, and cardiopulmonary bypass was at the discretion of the surgical team. Follow-up was not standardized but commonly included physical examination, complete serum biochemistry, and computerized tomography or ultrasound every 3 months in the first year and semiannually thereafter.