In the United Kingdom, there are approximately 1.3 million individuals diagnosed with atrial fibrillation (AF). This disease is associated with many serious complications, including heart failure and stroke. To prevent AF-related strokes, individuals with AF are typically prescribed blood thinners, such as warfarin, which have been used for stroke prevention in this population for over 60 years. In addition to stroke prevention, warfarin has been studied as a potential anti-cancer drug, with a recent observational study reporting a 16% decreased risk of cancer among warfarin users compared to non-users. However, not all studies have reported a consistent decreased risk. Over the past few years, the use of warfarin has been gradually replaced by the use of direct oral anticoagulants (DOACs), which were approved for stroke prevention in individuals with AF in 2011. Unlike warfarin, the use of DOACs does not require constant blood monitoring, though its safety, especially among vulnerable populations, is less established. Indeed, signals from clinical trials support a potential increased risk of certain cancers among DOAC users compared to placebo. While there have been several studies to date assessing the relationship between warfarin use and cancer incidence, no study has included DOAC users. Thus, using the Clinical Practice Research Datalink, we will investigate this possible link within a large group of patients with AF to address whether warfarin and DOACs are associated with cancer incidence.
Vitamin K antagonists (VKAs) and direct oral anticoagulants (DOACs) are two classes of anticoagulants, both used to prevent and treat complications of atrial fibrillation (AF). Both classes of drugs effectively reduce the risk of stroke, though they have different risk-benefit profiles. Indeed, use of VKAs requires careful monitoring through International Normalised Ratio (INR) testing in order to balance the anticoagulation benefits against the potential risk of haemorrhage. However, VKAs have been reported to have anti-cancer properties, with a recent study reporting a 16% decreased risk in incident cancer. On the other hand, use of DOACs does not require INR testing, as it has a more predictable pharmacokinetic profile, though its long-term risks are unknown. Indeed, signals from clinical trials indicate that use of DOACs may be associated with certain cancers. While there have been several studies investigating the association between warfarin and cancer incidence, no study has been conducted to specifically investigate this risk among DOAC users.
Thus, we will assemble a cohort of approximately 55,000 patients newly-diagnosed with AF between August 1, 2011 and December 31, 2017, with follow up until December 31, 2018. The cohort will start in 2011 as this is the year that the first DOAC was approved for stroke prevention in the UK. Use of anticoagulants will be modelled as a time-varying variable, allowing for a 1-year lag period for latency. Time-dependent Cox proportional hazards models will be used to estimate hazard ratios with 95% confidence intervals of cancer associated with use of VKAs and DOACs, separately, compared with non-use of either study drug. Secondary analyses will assess whether risk varies by duration of use, time since initiation, by cancer subtype, or according to specific VKAs (warfarin, acenocoumarol and phenindione) or DOACs (dabigatran, rivaroxaban, apixaban and edoxaban).
The primary outcome of interest for this study is an incident diagnosis of any cancer, identified using Read codes (Appendix I). Secondary outcomes, which will be considered hypothesis generating, are the most commonly diagnosed cancers in the UK, including breast, prostate, colorectal and lung,1 as well as those with numerical imbalances identified in clinical trials, including bladder and pancreatic cancer.
Samy Suissa - Chief Investigator - Sir Mortimer B Davis Jewish General Hospital
Laurent Azoulay - Corresponding Applicant - McGill University
Christel Renoux - Collaborator - McGill University
Devin Abrahami - Collaborator - McGill University
Hui Yin - Collaborator - Sir Mortimer B Davis Jewish General Hospital
Jean-Pascal Fournier - Collaborator - University of Nantes