Wide Variation in Duration of Anticoagulation after Catheter-Related Thrombosis: A Venous Thromboembolism Network US Multicenter Retrospective Cohort Study

Background: Consensus guidelines recommend anticoagulation for 3 months in patients with upper extremity catheter-related thrombosis (CRT); however, recommendations are primarily extrapolated from data on provoked lower extremity thrombosis. Many patients with hematologic malignancy are treated with shorter durations of anticoagulation due to thrombocytopenia and other bleeding concerns. It is not known if duration of anticoagulation in patients with hematologic malignancy patients and CRT predicts risk of recurrence or hemorrhage.

Methods: We performed a multicenter retrospective cohort study at five institutions through the Venous thromboEmbolism Network US (VENUS). Patients with hematologic malignancies with documented deep vein thrombosis associated with a catheter from 1/1/2010-12/31/2016 were identified using ICD-9 and ICD-10 diagnostic codes. Demographic information, management of the CRT, duration of anticoagulation, recurrent thrombosis, and bleeding during anticoagulation were extracted from the electronic medical record. Descriptive statistics will be used for demographic variables, disease and thrombosis characteristics, treatment type, duration of anticoagulation, recurrent thrombosis and bleeding outcomes. Non-parametric Kruskal Wallis were used to evaluate platelet count between treatment groups. Mann-Whitney U test was used to determine if duration of anticoagulation was associated with thrombosis recurrence.

Results: Of the 379 patients included in the analysis, 151 (40%) had lymphoma, 143 (38%) had leukemia, 74 (20%) had plasma cell disorder, and 11 (3%) had myelodysplastic syndrome or a myeloproliferative neoplasm. 214 (56%) were male. Median follow-up duration was 274 days, and 134 (35%) patients died within the follow-up period. PICCs accounted for 254 (67%) of the CRT events, 74 (20%) of cases had tunneled catheters, and 50 (13%) had Port-a-Caths. The most proximal involvement of the thrombosis were as follows: 37 (10%) brachial vein, 81 (21%) axillary vein, 149 (39%) subclavian vein, 90 (24%) internal jugular vein, 5 (1%) brachiocephalic vein, and 14 (4%) superior vena cava. 252 (66%) patients were treated with anticoagulation and removal of the catheter, but 50 (13%) were treated with anticoagulation alone, 60 (16%) were treated with removal of the catheter only, and 17 (4%) did not have treatment for the CRT. Median platelet count at diagnosis differed between the treatment groups [anticoagulation: 145,000 (range 27,000-443,000); anticoagulation and catheter removal: 130,000 (6,000-963,000); catheter removal only: 42,000 (5,000-251,000); no treatment: 39,000 (6,000-423,000); p<0.0001]. For the 302 patients treated with anticoagulation, 229 (76%) were initially treated with low-molecular weight heparin, 31 (10%) were treated with unfractionated heparin, and 30 (10%) were bridged to warfarin. Median duration of anticoagulation therapy for the initial CRT was 56.5 days (range 1- 1,108 days). Among the 340 patients who had their catheter removed, 245 (72%) were removed due to the CRT, 56 (16%) due to completion of cancer therapy, 22 (6%) due to infection, and 17 (5%) for other reasons. Fifty-six patients (15%) experienced a recurrent VTE, and 7 (2%) patients experienced a second recurrent VTE event. Thirty-eight (10%) patients experienced bleeding during anticoagulation of which 19 (5%) were major hemorrhages. Duration of first anticoagulation did not differ between patients with and without recurrent thrombosis [with recurrence: median 67 days (range: 4-1,108), without recurrence: 54 (range: 1-453, p=0.32)].

Conclusions: Significant variability remains in treatment of CRT in patients with hematologic malignancy. Despite the guidelines, many catheters are removed due to the thrombosis. Our results suggest that duration of anticoagulation is not associated with VTE recurrence.