In this French nationwide cohort of 683,184 patients with incident treated venous thromboembolism, DOAC use increased substantially in patients with active cancer (15.3% in 2016-2018 to 32.6% in 2021-2022) but remained lower than in patients without active cancer and varied across clinical and care-setting characteristics.
Key Findings
Results
Among 683,184 patients with incident treated venous thromboembolism in France (2016-2022), 13.5% had active cancer.
91,954 patients (13.5%) had active cancer and 591,230 (86.5%) had VTE not associated with active cancer.
Median (IQR) age was 67 (53-77) years; 347,421 patients (51.0%) were female.
Data came from the French National Health Data System, a claims database covering about 99% of the population; adults without recorded DOAC contraindications and with a first treated VTE between January 1, 2016, and December 31, 2022 were included.
Results
DOAC use for VTE roughly doubled among patients with active cancer between 2016-2018 and 2021-2022.
DOAC use rose from 15.3% (95% CI, 15.0%-15.7%) in 2016-2018 to 32.6% (95% CI, 32.1%-33.2%) in 2021-2022.
The primary outcome was DOAC initiation within 30 days after the index VTE event.
Results
DOAC use was also increasing among patients without active cancer and remained much higher than in patients with active cancer.
Use rose from 63.5% (95% CI, 63.3%-63.7%) in 2016-2018 to 81.1% (95% CI, 80.9%-81.2%) in 2021-2022.
In 2021-2022, DOAC use was 32.6% in active cancer versus 81.1% without active cancer.
Results
In patients with active cancer, DOAC initiation increased beyond expected secular trends after 2018.
Bayesian structural time series models estimated a relative increase of 20.5% (95% credible interval, 13.0%-27.8%) in 2019-2020.
The relative increase was 41.3% (95% credible interval, 32.9%-50.2%) in 2021-2022.
Periods compared were 2016-2018, 2019-2020, and 2021-2022.
Results
Among patients with active cancer, several clinical characteristics were associated with less frequent DOAC initiation.
Lower DOAC initiation was associated with comorbidities, history of bleeding, gastrointestinal or gynecologic cancer, active brain metastases, and systemic treatment.
Identified using multivariable logistic regression among patients with active cancer.
Results
Initial prescriber setting was associated with DOAC initiation in active cancer.
DOAC initiation was less frequent when the initial prescriber was a hospital-based physician rather than a general practitioner.
Conclusions
The authors interpret the lower DOAC uptake in cancer-associated VTE as possibly reflecting delayed implementation of recent evidence and the need to individualize anticoagulant choice in complex situations.
DOAC use remained lower in active cancer than in patients without active cancer despite substantial increases.
Use varied across clinical and care-setting characteristics.
What This Means
This study used French national health insurance data covering about 99% of the population to look at how blood-thinning treatment for blood clots (venous thromboembolism) changed from 2016 to 2022. It focused on direct oral anticoagulants (DOACs), pills that are now recommended for some patients with cancer-related clots. Of 683,184 people treated for a first clot, about 13.5% had active cancer.
Among people with active cancer, DOAC use more than doubled, from about 15% in 2016-2018 to about 33% in 2021-2022. The rise was larger than would be expected from general trends alone. Among people without active cancer, DOAC use was already high and rose from about 64% to 81%. So cancer patients were still much less likely to receive DOACs. Within the cancer group, DOACs were used less often in people with other health conditions, a past history of bleeding, gastrointestinal or gynecologic cancers, active brain metastases, or those receiving systemic cancer treatment. They were also used less often when the first prescriber was a hospital doctor rather than a general practitioner.
This research suggests that guideline-recommended DOAC use for cancer-associated clots has been spreading but is still uneven. The lower use may partly reflect slow adoption of new evidence and partly appropriate caution in complex patients. As an observational claims-based study, it describes patterns of prescribing and cannot say which treatment choices were best for individual patients.
Varnier R, Bettega F, Moskal A, Mainbourg S, Wespiser M, Rousseau A, et al.. (2026). Trends in Direct Oral Anticoagulant Use for Cancer-Associated Venous Thromboembolism.. JAMA network open. https://doi.org/10.1001/jamanetworkopen.2026.40377