1. Introduction
Cancer-associated thrombosis (CAT) is a common and highly lethal complication in cancer patients, garnering significant attention in recent years. CAT not only severely affects patient prognosis but also significantly reduces quality of life, becoming the second leading cause of death in cancer patients [1][2]. The mechanisms underlying CAT are complex, involving abnormal expression of procoagulant factors by tumor cells, vascular endothelial damage, and a hypercoagulable state of the blood [3][4]. Different types of cancers, such as pancreatic cancer, brain cancer, gastric cancer, and hematological malignancies, carry a higher risk of thrombosis [5][1]. Additionally, tumor progression, chemotherapy, and targeted therapies can further increase the risk of thrombosis [6][7].
Although low molecular weight heparin (LMWH) has long been considered the standard treatment for CAT [8][9], direct oral anticoagulants (DOACs) have been increasingly introduced into the treatment and prevention of CAT due to their oral convenience, fewer drug interactions, and lack of need for routine monitoring [10][11]. Multiple randomized controlled trials have shown that DOACs are non-inferior to LMWH in preventing recurrence of tumor-related venous thromboembolism, although some patients, particularly those with gastrointestinal and genitourinary malignancies, have an increased risk of bleeding [12][10][9]. Therefore, the clinical application of DOACs must fully consider the type of cancer, bleeding risk, drug interactions, and individual patient characteristics, emphasizing precision medicine [11][13].
Traditional anticoagulant therapy has limitations, including uncertain efficacy and a higher risk of bleeding, especially in patients with thrombocytopenia or liver dysfunction [14][15]. Furthermore, cancer patients often experience complex drug interactions, particularly between DOACs and anticancer drugs, which may affect the efficacy and safety of anticoagulation [16][17]. Consequently, individualized risk assessment and treatment strategies are urgently needed in the management of CAT [18][19]. Various risk assessment tools, such as the Khorana score, COMPASS-CAT, Modified Khorana, and machine learning models, have shown predictive value in risk stratification for CAT patients, but issues such as inaccurate prognosis and lack of widespread validation remain [20][21][22]. Future efforts should focus on optimizing risk prediction models by integrating patient clinical characteristics, tumor type, and biomarkers to guide the selection and duration of anticoagulant therapy [23][24].
In the choice and dosing of anticoagulant medications, the rational use of LMWH, DOACs, and warfarin, among others, is crucial based on the specific condition of the patient [25][26][11]. Some studies have shown that DOACs exhibit better adherence and improvements in quality of life in certain patients [27][28]. Additionally, perioperative anticoagulation management, monitoring and adjustment of anticoagulant drugs, and the safety of drug use in special populations, such as elderly patients and those with impaired liver or kidney function, require careful consideration [29][30][31]. Moreover, with the development of artificial intelligence, big data, and precision medicine, machine learning-based CAT prediction models and personalized medication regimens are becoming research hotspots [19][32].
In summary, the anticoagulant treatment of cancer-associated thrombosis is moving towards individualization and precision. By integrating clinical risk assessment tools, molecular biomarkers, and novel anticoagulant drugs, personalized anticoagulation strategies can be developed to maximize therapeutic benefits, minimize bleeding risks, and improve patient outcomes and quality of life. Future research should further refine risk assessment models, clarify the application of DOACs in specific cancer subtypes and complex clinical scenarios, and promote multidisciplinary collaboration to establish a systematic CAT management system [33][34].
