Abstract
INTRODUCTION: Radical nephrectomy (RN) with venous tumour thrombectomy (VTT) carries a significant morbidity and mortality risk. Examination of a contemporary single-institution series permits the development of a management algorithm and an audit its results. We report outcomes following the use of intraoperative colour Doppler ultrasound and our surgical pathway. METHODS: We retrospectively reviewed the records of all patients who underwent RN with VTT for kidney cancer between January 1, 2013 and October 1, 2016. Surgical complications, postoperative complications (Clavien-Dindo classification ≥3), 90-day readmission rates, and outcomes are reported. Multivariate linear regression, logistic regression, and Cox proportional hazard modelling were used to identify associations. RESULTS: Fifty-eight patients underwent RN with VTT. Of these, 26 (45%) patients had Mayo Clinic level III or IV thrombus and nineteen required venovenous/cardiopulmonary bypass. Three patients required patch grafting. The median length of hospital stay was eight days and there were 20 major complications. The 30-day readmission rate was 21% and the 90-day mortality rate was 8.9%. In multivariate analysis, low serum albumin and age-adjusted Charlson comorbidity score predicted length of stay. Increased intraoperative blood loss was significantly associated with increasing body mass index, serum creatinine, tumour thrombus level, and a history of significant weight loss >9.1kg. Low serum hematocrit predicted 90-day mortality. CONCLUSIONS: Intraoperative colour Doppler ultrasound is a useful tool and can facilitate caval preservation. Caval grafting can be avoided in most cases. Venovenous bypass can be avoided in many level III cases. Early therapeutic anticoagulation should be instituted with caution.