Is imperative partial nephrectomy feasible for kidney cancer with venous thrombus involvement? Outcomes of 42 cases and matched pair analysis with a large radical nephrectomy cohort

Background: Radical nephrectomy (RN) with/without (±) thrombus excision (ThE) is the undisputed standard treatment for kidney cancer (KC) with renal or caval thrombus (Th). However, partial nephrectomy (PN) ± ThE may be considered in rare cases due to imperative (I) indications. Objective: To evaluate the efficacy of IPN ± ThE and to compare it with RN ± ThE for KC with Th. Design, setting, and participants: Records of 2,549 patients undergoing surgery for KC with Th at 24 institutions between 1971 and 2014 were retrospectively reviewed. Outcome measurements and statistical analysis: Primary outcomes were overall survival (OS) and cancer specific survival (CSS), renal function variation after surgery and complications. Secondary outcomes were predictors of OS and CSS for IPN cases. To reduce bias IPN group was matched with RN using a propensity score with greedy algorithm on the basis of age, gender, tumor size, TNM, and histology. Results and limitations: Forty-two patients underwent IPN ± Th. All thrombi were ≥level I; 5 patients experienced Clavien ≥ 3 complications with 2 complications-related deaths. At 27.3 (interquartile range: 7.1–47.7) months OS and CSS were 54.8% and 78.6%, respectively whereas at 9.7 (interquartile range: 1.4–43.7) months eGFR change was −17.3 ± 27.0 ml/min. On univariate analysis tumour size, preoperative eGFR, transfusions, hospital stay, high serum creatinine, operating time, complications, lymphadenectomy, and metastases related to an increased risk of death. After matching (n = 38 per arm) no significant differences were present except for tumor necrosis (IPN = 39.5%; 15.8%; P = 0.01), thrombus level (P = 0.02), so as for operating time (P = 0.27), perioperative transfusions (P = 0.74) and complications (P = 0.35). A 5-year OS and CSS for IPN were 57.9% and 73.7%, respectively with no significant differences with RN (OS = 63.2, P = 0.611; CSS = 68.4, P>0.99). After 14.9 months creatinine and eGFR changes were (+0.4 ± 0.6 mg/dl and −23.2 ± 37.3 ml/min; P = 0.2879). Conclusions: In selected cases due to imperative indications PN ± ThE is a complex procedure and may be an alternative to RN ± ThE for KC with Th yielding noninferior oncological outcomes, functional outcomes, and complications. Further studies are needed to determine the role of PN ± ThE for KC with Th.