Research Group "Cancer, Haemostasis and Angiogenesis", INSERM U938, Centre de Recherche Saint-Antoine, Institut Universitaire de Cancérologie, Faculty of Medicine, Sorbonne University, Paris, France; Service d'Hématologie Biologique Hôpital Tenon, Hôpitaux Universitaires de l'Est Parisien, Assistance Publique Hôpitaux de Paris, Paris, France; Department of Hematology and Cell Therapy, Saint Antoine Hospital, Hôpitaux Universitaires de l'Est Parisien, Assistance Publique Hôpitaux de Paris, Sorbonne University, Paris, France.
Internal Medicine Department, Hôpital Louis Mourier, APHP, Colombes, Inserm UMR_S1140, Université Paris-Diderot Paris7, Paris, France.
Thromb Res. 2020 Jul;191 Suppl 1:S50-S57. doi: 10.1016/S0049-3848(20)30397-2.
A B S T R A C T Important progress has been made in the development of risk assessment models (RAM) for the identification of outpatients on anticancer treatment at risk of venous thromboembolism (VTE). Since the breakthrough publication of the original Khorana risk score (KRS) more than 10 years ago, a new generation of KRS-based scores have been developed, including the Vienna Cancer and Thrombosis Study, PROTECHT, CONKO, ONCOTEV, TicOnco and the CATS/MICA score. Among these the CATS/MICA score showed that a simplified score composed of only two calibrated predictors, the type of cancer and the D-dimer levels, offers a user-friendly tool for the evaluation of cancer-associated thrombosis (CAT) risk. The COMPASS-CAT score is the first that introduced a more synthetic approach of risk evaluation by combining cancer-related predictors with patient comorbidity in a score which is designed for the types of cancer frequently seen in the community (i.e. breast, lung colon or ovarian cancers) and has been externally validated in independent studies. The Throly score is registered as part of the same group as it has a similar structure to the COMPASS-CAT score and is applicable in patients with lymphoma. The incorporation of specific biomarkers of hypercoagulability to the RAM for CAT offers the possibility to perform a precision medicine approach in the prevention of CAT. The improvement of RAM for CAT with artificial intelligence methodologies and deep learning techniques is the challenge in the near future.
在开发用于识别接受抗肿瘤治疗的门诊患者静脉血栓栓塞 (VTE) 风险的风险评估模型 (RAM) 方面已经取得了重要进展。自 10 多年前原始 Khorana 风险评分 (KRS) 的突破性发表以来，已经开发出了新一代基于 KRS 的评分，包括维也纳癌症和血栓形成研究、PROTECHT、CONKO、ONCOTEV、TicOnco 和 CATS/MICA 评分。其中，CATS/MICA 评分表明，仅由两个校准预测因子（癌症类型和 D-二聚体水平）组成的简化评分提供了一种用于评估癌症相关血栓形成 (CAT) 风险的简便工具。COMPASS-CAT 评分是第一个通过将癌症相关预测因子与患者合并症相结合，在一个针对社区中常见癌症类型（即乳腺癌、肺癌、结肠癌或卵巢癌）设计的评分中引入更综合的风险评估方法的评分，并在独立研究中得到了外部验证。Throly 评分是作为同一组的一部分注册的，因为它与 COMPASS-CAT 评分具有相似的结构，适用于淋巴瘤患者。将特定的高凝状态生物标志物纳入 CAT 的 RAM 提供了在 CAT 预防中实施精准医学方法的可能性。在不久的将来，人工智能方法和深度学习技术对 CAT 的 RAM 的改进是一个挑战。
