Perioperative Cerebrospinal Fluid Drain Placement Does Not Increase Venous Thromboembolism Risk After Thoracic and Fenestrated Endovascular Aortic Repair

Abstract

Venous thromboembolism (VTE) incidence after thoracic and fenestrated endovascular aortic repair (TEVAR/FEVAR) is high (up to 6-7%) relative to other vascular procedures; however, the etiology for this discrepancy remains unknown. Notably, patients undergoing TEVAR/FEVAR commonly receive cerebrospinal fluid drains (CSFDs) for neuroprotection, requiring interruption of perioperative anticoagulation and prolonged immobility. We hypothesized that CSFDs are a risk factor for VTE after TEVAR/FEVAR. Consecutive TEVAR/FEVAR patients at a single center were reviewed (2011-2020). CSFDs were placed based on surgeon preference preoperatively or for spinal cord ischemia (SCI) rescue therapy postoperatively. The primary end-point was VTE occurrence, defined as any new deep venous thrombosis (DVT) or pulmonary embolism (PE) confirmed on imaging within 30-days postoperatively. Routine postoperative VTE screening was not performed. Patients with and without VTE, and subjects with and without CSFDs were compared. Logistic regression was used to explore associations between VTE incidence and CSFD exposure. 897 patients underwent TEVAR/FEVAR and 43% (n=387) received a CSFD at some point during their care (preoperative: 94%[n=365/387]; post-operative SCI rescue therapy: 6%[n=22/387]). CSFD patients were more likely to have previous aortic surgery (44% vs. 37%; p=.028) and received more postoperative blood products (780 vs. 405mL; p=.005). The overall VTE incidence was 2.2% (n=20). 70% (14) patients with VTE had DVT, 50% (10) had PE, and 20% (4) had DVT and PE. Among TEVAR/FEVAR patients with VTE, 65% (n=13) were symptomatic. Most VTEs (90%, n=18) were identified in-hospital and the median time to diagnosis was 12.5 (IQR 7.5-18) days postoperatively. Patients with VTE were more likely to have non-elective surgery (95% vs. 41%; p<.001), had higher ASA classification (4.1 vs. 3.7; p<.001), required longer ICU admission (24 vs. 12-days; p<.001), and received more blood products (1386 vs. 559mL; p<.001). VTE incidence was 1.8% in CSFD patients compared to 3.5% in non-CSFD patients (OR 0.70 [95% CI 0.28-1.78, p = .300). However, patients receiving CSFDs postoperatively for SCI rescue therapy had significantly greater VTE incidence (9.1% vs. 1.1%; p=.044). CSFD placement was not associated with an increased risk of VTE in patients undergoing TEVAR/FEVAR. VTE risk was greater in patients undergoing nonelective surgery and those with complicated perioperative courses. VTE risk was greater in patients receiving therapeutic CSFDs compared to prophylactic CSFDs, highlighting the importance of careful patient selection for prophylactic CSFD placement.