PD10-05: Perioperative Blood Transfusion is Associated with Increased Risk of Venous Thromboembolism after Radical Cystectomy

Amanda Myers*, Jacksonville, FL, Igor Frank, Paras H. Shah, Robert F. Tarrell, Rochester, MN, Giovanni A Gonzalez, Jacksonville, FL, R. Jeffrey Karnes, Houston Thompson, Matthew K. Tollefson, Stephen A. Boorjian, Rochester, MN, Timothy D. Lyon, Jacksonville, FL

Introduction: Venous thromboembolism (VTE) is increasingly used as a quality indicator after radical cystectomy (RC). However, the role of perioperative blood transfusion (PBT) in VTE risk adjustment remains unclear. While transfusion has been associated with an increased risk of VTE in administrative datasets, these analyses are limited by inability to adjust for several risk factors for VTE including cancer stage and receipt of neoadjuvant chemotherapy which may confound the observed associations. The objective of this study was to examine whether perioperative blood transfusion is associated with VTE following RC after adjusting for both patient- and disease-related factors.

Methods: Patients who underwent RC for bladder cancer from 1980-2020 were identified in the Mayo Clinic cystectomy registry. PBT was defined as intraoperative and/or postoperative blood transfusion during the initial hospital stay. The primary outcome was VTE within 90 days of RC, inclusive of either deep venous thrombosis or pulmonary embolism. Perioperative VTE prophylaxis varied according to institutional standard at the time, but generally consisted of sequential compression devices and subcutaneous heparin while inpatient. In April 2013, we began using intraoperative tranexamic acid, and in May 2014, we began prescribing outpatient extended duration VTE prophylaxis with enoxaparin in all patients without contraindications. Associations between clinicopathologic variables and 90-day VTE were assessed using multivariable logistic regression. PBT was analyzed as both a categorical and a continuous variable.

Results: We identified 3958 patients with RC, of whom 162 (4%) experienced a VTE within 90 days of RC. 2244 patients (57%) received a PBT, with a median of 1 (range: 0-69, IQR: 0-3) unit transfused. Median follow up among survivors was 6.7 years (IQR 2.4-12.7). On multivariable analysis adjusting for year of surgery, BMI, robotic approach, neoadjuvant chemotherapy, diversion type, pathologic tumor stage and pathologic nodal stage, the receipt of PBT remained significantly associated with an increased risk of VTE (adjusted OR 1.46, 95% CI 1.01-2.09, p=0.04). When further analyzed as a continuous variable, each unit of blood was associated with 6% higher odds of a VTE (adjusted OR 1.06, 95% 1.01-1.10, p=0.01).

Conclusions: PBT was independently associated with VTE within 90 days of RC after adjusting for disease-related factors. Each unit transfused was associated with higher odds of a VTE. Continued effort to limit blood transfusion in these patients is warranted.

Source of Funding: none