Contributions
Abstract: S441
Type: Oral Presentation
Presentation during EHA22: On Saturday, June 24, 2017 from 11:30 - 11:45
Location: Room N103
Background
Cancer patients have a higher risk of venous thromboembolism (VTE) which conveys a higher subsequent mortality risk; conversely, they also have a higher risk for bleeding due to many factors including abnormal tumor anatomy and the use of chemotherapy agents with the associated risk for thrombocytopenia. However, the consequences of a recurrent VTE or a major bleeding (MB) event might be different in terms of mortality. As a result, the risk of VTE recurrence or a MB event might bear different weights. A previous systematic review has suggested that the case fatality rates of VTE recurrence and MB are similar. However, heterogeneity in study design, outcomes and in particular the types of populations included, limited the interpretation and applicability of the results.
Aims
To estimate case fatality rates of VTE recurrence and MB, as well as the case fatality rate-ratio for MB and VTE recurrence in cancer patients developing a VTE treated with anticoagulants.
Methods
We conducted a retrospective population-based cohort study in Ontario, Canada using de-identified linked administrative healthcare databases housed at the Institute for Clinical Evalutive Sciences (ICES). We included patients over 65 years of age with a diagnosis of cancer defined using provincial, ICD-9 and ICD-10 codes for major malignancies and who developed a VTE event within 6 months of the initial cancer diagnosis. VTE was identified through a previously validated algorithm using a combination of diagnostic codes for deep vein thrombosis (DVT) and pulmonary embolism (PE) and codes identifying diagnostic procedures for VTE (i.e. ultrasound, CT pulmonary angiography, lung scintigraphy) within 7 days of each other. Recurrent VTE and MB events were assessed within 180 days from the index date. MB was identified using a previously validated algorithm and included upper and lower gastrointestinal and intracranial bleeding events. Treatment was classified based on the first available prescription within 7 days of the index VTE. We estimated mortality within 7 days of the VTE recurrence or MB events using an unadjusted Cox proportional hazards model and competing risk analysis. Ratios of the mortality for MB compared to VTE recurrence were calculated and 95% confidence intervals were estimated using non-parametric models.
Results
Between 2004 and 2014 there were 6967 VTE events identified in cancer patients over 65 years of age and treated with an anticoagulant. Mean age was 75 years, and 47.6% patients were women. Of all patients, 59.9% received prescriptions for LMWH alone, 15.3% for LMWH followed by warfarin, 22.1% for warfarin and 2.7% for rivaroxaban. At 180 days after the index VTE event there were 235 (3%) MB events and 1184 (17%) VTE recurrences. Within 7 days of the outcome event there were 26 (11%) deaths after MB and 6 (0.5%) after VTE. The mortality ratio for MB versus VTE was 21.8 (95% CI 9-53). In exploratory analyses we did not find differences according to type of anticoagulant prescription.
Conclusion
In cancer patients 65 years or older treated with anticoagulants for a VTE, the 7 days mortality after a MB event is at least 9 times higher than after a VTE recurrence, although the estimate is imprecise. This data should be taken into consideration when designing studies and interventions involving anticoagulant therapy in this population.
Session topic: 34. Thrombosis and vascular biology
Abstract: S441
Type: Oral Presentation
Presentation during EHA22: On Saturday, June 24, 2017 from 11:30 - 11:45
Location: Room N103
Background
Cancer patients have a higher risk of venous thromboembolism (VTE) which conveys a higher subsequent mortality risk; conversely, they also have a higher risk for bleeding due to many factors including abnormal tumor anatomy and the use of chemotherapy agents with the associated risk for thrombocytopenia. However, the consequences of a recurrent VTE or a major bleeding (MB) event might be different in terms of mortality. As a result, the risk of VTE recurrence or a MB event might bear different weights. A previous systematic review has suggested that the case fatality rates of VTE recurrence and MB are similar. However, heterogeneity in study design, outcomes and in particular the types of populations included, limited the interpretation and applicability of the results.
Aims
To estimate case fatality rates of VTE recurrence and MB, as well as the case fatality rate-ratio for MB and VTE recurrence in cancer patients developing a VTE treated with anticoagulants.
Methods
We conducted a retrospective population-based cohort study in Ontario, Canada using de-identified linked administrative healthcare databases housed at the Institute for Clinical Evalutive Sciences (ICES). We included patients over 65 years of age with a diagnosis of cancer defined using provincial, ICD-9 and ICD-10 codes for major malignancies and who developed a VTE event within 6 months of the initial cancer diagnosis. VTE was identified through a previously validated algorithm using a combination of diagnostic codes for deep vein thrombosis (DVT) and pulmonary embolism (PE) and codes identifying diagnostic procedures for VTE (i.e. ultrasound, CT pulmonary angiography, lung scintigraphy) within 7 days of each other. Recurrent VTE and MB events were assessed within 180 days from the index date. MB was identified using a previously validated algorithm and included upper and lower gastrointestinal and intracranial bleeding events. Treatment was classified based on the first available prescription within 7 days of the index VTE. We estimated mortality within 7 days of the VTE recurrence or MB events using an unadjusted Cox proportional hazards model and competing risk analysis. Ratios of the mortality for MB compared to VTE recurrence were calculated and 95% confidence intervals were estimated using non-parametric models.
Results
Between 2004 and 2014 there were 6967 VTE events identified in cancer patients over 65 years of age and treated with an anticoagulant. Mean age was 75 years, and 47.6% patients were women. Of all patients, 59.9% received prescriptions for LMWH alone, 15.3% for LMWH followed by warfarin, 22.1% for warfarin and 2.7% for rivaroxaban. At 180 days after the index VTE event there were 235 (3%) MB events and 1184 (17%) VTE recurrences. Within 7 days of the outcome event there were 26 (11%) deaths after MB and 6 (0.5%) after VTE. The mortality ratio for MB versus VTE was 21.8 (95% CI 9-53). In exploratory analyses we did not find differences according to type of anticoagulant prescription.
Conclusion
In cancer patients 65 years or older treated with anticoagulants for a VTE, the 7 days mortality after a MB event is at least 9 times higher than after a VTE recurrence, although the estimate is imprecise. This data should be taken into consideration when designing studies and interventions involving anticoagulant therapy in this population.
Session topic: 34. Thrombosis and vascular biology