Clinically relevant bleeding according to location of metastases in cancer-associated thrombosis.

Abstract

Patients with cancer-associated thrombosis (CAT) face a heightened risk of clinically relevant bleeding (CRB). However, the relationship between these risks and the location of metastasis remains unclear. A single-center, non-interventional study of consecutive patients was conducted between 2007 and 2022, involving patients with CAT (n = 1,277). Our primary objectives were: 1) To estimate the rate of CRB based on metastasis location. 2) To assess the long-term risk of CRB. Over a median follow-up of 14.2 months, 144 CRB were observed. The most frequent cancers were colon (19.2 %), lung (16.1 %), and breast (12.4 %). Fifty-two per-cent had metastases. Compared to patients without metastases, metastasis locations associated with CRB were lung metastases (rate 10.0 per 100 patients/year; 95 % confidence interval [CI] 6.6 to 14.6; risk ratio [RR]: 2; 95 % CI: 1.3-3.1), liver metastases (rate 10.1 per 100 patients/year; 95 % CI 6.5-14.9; RR: 2.0; 95 % CI: 1.3-3.2) and adenopathy metastases (rate 11.9 per 100 patients/year; 95 % CI 6.8-19.3; RR: 2.4; 95 % CI: 1.4-4.1). Multivariate analysis of variables associated to CRB at long term follow-up included ECOG > 1 (HR 3.0, 95 % CI 1.7-5.3), bladder cancer (HR 2.5, 95 % CI 1.3-4.8), prostate cancer (HR 2.2, 95 % CI 1.1-4.8) and lung metastases (HR 2.1, 95 % CI 1.3-3.6). The bleeding rate seem differ depending on the location of metastasis. Additionally, lung metastasis was found to be associated with a long-term risk of CRB. These findings may justify the use of low dose of anticoagulation, although clinical trials need to demonstrate the efficacy and safety of this strategy.