Comparative safety outcomes of three venous access routes for totally implantable venous access port placement and exploratory assessment of pre-removal catheter-related thrombosis in breast cancer patients.

Objective: To compare the safety outcomes of totally implantable venous access port (TIVAP) implantation through the internal jugular vein (IJV), subclavian vein (SCV), and axillary vein (AxV) in patients with breast cancer, and to explore a pre-removal thrombosis risk-stratification framework integrating D-dimer levels and vascular ultrasonography findings. Methods: Data from 475 breast cancer patients who underwent TIVAP implantation were retrospectively reviewed. Patients were classified according to implantation route: IJV (n = 161), SCV (n = 138), and AxV (n = 176). Baseline characteristics, procedure variables, postoperative complications, and Clavien-Dindo grades were analyzed. Among these patients, 234 underwent both pre-removal D-dimer testing and vascular ultrasonography, and 18 asymptomatic thrombotic events were identified. Firth penalized logistic regression was used to explore factors associated with thrombosis. ROC curve analysis was performed to evaluate the discriminatory ability of D-dimer, and the optimal cutoff was determined using the Youden index. Patients were categorized into four scenarios according to D-dimer and ultrasonography findings. Results: Baseline characteristics were comparable among the three groups. The complication rate was higher in the SCV group than in the IJV and AxV groups (16.67% vs. 7.45% vs. 5.68%, P = 0.002). Catheter fracture and pinch-off syndrome occurred exclusively in patients receiving SCV access. Firth penalized logistic regression showed that higher D-dimer levels and longer catheter dwell time were associated with asymptomatic catheter-related thrombosis. D-dimer showed moderate discrimination (AUC = 0.721). At the ROC-derived cutoff of 0.251 μg/mL, sensitivity was 77.78%, specificity was 76.85%, LR+ was 3.36, and LR- was 0.29. Integration of D-dimer and ultrasonography yielded four exploratory pre-removal profiles, including 166 double-negative, 4 D-dimer-negative/ultrasound-positive, 50 D-dimer-positive/ultrasound-negative, and 14 double-positive patients. Conclusion: Compared with the SCV approach, the AxV approach showed a lower observed complication rate in this retrospective cohort, whereas no statistically significant difference in overall complications was observed between the AxV and IJV groups. D-dimer alone was insufficient to exclude thrombosis. The combined D-dimer/ultrasonography framework should be regarded as exploratory and requires prospective validation before its potential clinical application can be considered.