J Thromb Haemost. 2026 Sep 7:S1538-7836(26)00557-X. doi: 10.1016/j.jtha.2026.08.032. Online ahead of print.
ABSTRACT
BACKGROUND: Upper-extremity deep vein thrombosis (UEDVT) is increasingly diagnosed because of the widespread use of venous catheters, particularly in patients with cancer. However, its clinical course and the optimal duration of anticoagulant therapy remain uncertain.
METHODS: We analyzed 4,979 patients with symptomatic isolated UEDVT enrolled in the RIETE registry, including 1,383 patients with cancer and catheter-associated UEDVT, 599 with cancer and non-catheter-associated UEDVT, 921 without cancer and catheter-associated UEDVT, and 2,076 without cancer and non-catheter-associated UEDVT. Outcomes included recurrent deep vein thrombosis, symptomatic pulmonary embolism, major bleeding, and all-cause mortality during anticoagulation and after treatment discontinuation. Competing-risk regression was used for thrombotic and bleeding outcomes, and Cox regression for mortality.
RESULTS: Catheter-associated UEDVT was more frequent among patients with cancer than among those without cancer (70% vs. 35%). Among patients with cancer, catheter-associated UEDVT was associated with lower all-cause mortality than non-catheter associated UEDVT (22.6 vs. 55.9 deaths per 100 patient-years; adjusted hazard ratio [aHR] 0.39; 95%CI 0.31-0.49) and lower recurrent VTE rates (6.27 vs. 9.60 per 100 patient-years). Among patients without cancer, catheter-associated UEDVT was associated with higher mortality (10.7 vs. 4.32 deaths per 100 patient-years) and major bleeding (2.49 vs. 0.61 per 100 patient-years), whereas recurrent VTE rates were similar irrespective of catheter status. Across all four groups, recurrent VTE remained infrequent after anticoagulation discontinuation.
CONCLUSIONS: Clinical outcomes of UEDVT differ according to both cancer status and catheter association. The persistently low recurrence rates after anticoagulation discontinuation, particularly in catheter-associated UEDVT, support an individualized approach to anticoagulation duration.
PMID:42705547 | DOI:10.1016/j.jtha.2026.08.032