Clinical analysis of bleeding and thrombotic events in haematological-oncology patients with severe thrombocytopenia and a high risk of thrombosis

Abstract

Abstract Background: Haematological patients with severe thrombocytopenia and high thrombotic risk face challenges related to balancing bleeding and thrombosis risks. This study investigated factors associated with bleeding and thrombosis in high-risk haematological oncology patients with severe thrombocytopenia not receiving anticoagulant therapy and characterized their clinical features when both events occurred. Methods: A total of 446 haematological oncology patients with Caprini scores ≥ 5 were included from July 2022 to June 2023 at Mianyang City Central Hospital. Those not receiving prophylactic anticoagulants due to an admission platelet count < 50 × 109/L were studied. Patients were categorized into bleeding/nonbleeding and thrombotic/nonthrombotic groups on the basis of hospital course. Relevant clinical data were collected, and univariate/multivariate logistic regression was used to analyse the influencing factors. The platelet count at admission was assessed via ROC curves for thrombosis prediction. Results: In the bleeding group, higher proportions of patients with leukaemia, myeloid tumours, lung infections, and a central venous catheter (CVC) with two lumens were observed, along with shorter catheter durations, lower initial and minimum platelet counts during hospitalization, and prolonged plasminogen times (all P < 0.05). The thrombotic group had a greater thrombosis history, initial platelet count, use of two venous catheter lumens, parenteral nutrition, sedation, and autologous haematopoietic stem cell transplantation (Auto-HSCT), with a lower leukaemia proportion (P < 0.05). Logistic regression identified lymphoma type and minimum platelet count as bleeding protective factors and the Charlson Comorbidity Index (CCI) score as an independent risk factor. Thrombosis history, two venous catheter lumens, and sedation were risk factors for thrombosis. The median platelet count was lower at bleeding and thrombosis than at admission (P = 0.007). The platelet count at admission had predictive value for thrombosis, especially severe thrombocytopenia, with an AUC of 0.735 (95% CI 0.613–0.858, P = 0.003) and a cut-off value of 42.5 × 109/L. Conclusions: For haematological neoplasm patients with a high risk of venous thromboembolism (VTE), severe thrombocytopenia and high CCI scores, risk prevention and control of bleeding take precedence over thrombosis prophylaxis. Prophylactic anticoagulation is still recommended for patients with lymphoma assessed at high risk for VTE and with platelet counts of at least 42.5 × 109/L.

Keywords

• Haematological Neoplasms
• Thrombosis
• Bleeding
• Thrombocytopenia
• Risk factors
• Thrombohemorrhagic syndrome (TSH)