The results, which were published in the Annals of Internal Medicine, could help guide DOAC selection for high-risk groups with a prior history of peptic ulcer disease or major GI bleeding, said lead study authors Arnar Bragi Ingason, MD and Einar S. Björnsson, MD, PhD, in an email.
DOACs treat conditions such as atrial fibrillation, venous thromboembolism, and ischemic stroke and are known to cause GI bleeding. Previous studies have suggested that rivaroxaban poses a higher GI-bleeding risk than other DOACs.
These studies, which used large administrative databases, “had an inherent risk of selection bias due to insurance status, age, and comorbidities due to their origin from insurance/administrative databases. In addition, they lacked phenotypic details on GI bleeding events,” said Dr. Björnsson and Dr. Ingason, who are both of Landspitali University Hospital, Reykjavik, Iceland,
Daily dosage may exacerbate risk
Rivaroxaban is administered as a single daily dose, compared with apixaban’s and dabigatran’s twice-daily regimens. “We hypothesized that this may lead to a greater variance in drug plasma concentration, making these patients more susceptible to GI bleeding,” the lead authors said.
Using data from the Icelandic Medicine Registry, a national database of outpatient prescription information, they compared rates of GI bleeding among new users of apixaban, dabigatran, and rivaroxaban from 2014 to 2019. Overall, 5,868 patients receiving one of the DOACs took part in the study. Among these participants, 3,217 received rivaroxaban, 2,157 received apixaban, and 494 received dabigatran. The researchers used inverse probability weighting, Kaplan–Meier survival estimates, and Cox regression to compare GI bleeding.
Compared with dabigatran, rivaroxaban was associated with a 63%-104% higher overall risk for GI bleeding and 39%-95% higher risk for major GI bleeding. Rivaroxaban also had a 40%-42% higher overall risk for GI bleeding and 49%-50% higher risk for major GI bleeding, compared with apixaban.
The investigators were surprised by the low rate of upper GI bleeding for dabigatran, compared with the other two drugs. “However, these results must be interpreted in the context that the dabigatran group was relatively small,” said Dr. Björnsson and Dr. Ingason via email.
Overall, the study cohort was small, compared with previous registry studies.
Investigators also did not account for account for socioeconomic status or lifestyle factors, such as alcohol consumption or smoking. “However, because the cost of all DOACs is similar in Iceland, selection bias due to socioeconomic status is unlikely,” the investigators reported in their paper. “We are currently working on comparing the rates of thromboembolisms and overall major bleeding events between the drugs,” the lead authors said.
Clinicians should consider location of bleeding
Though retrospective, the study by Ingason et. al. “is likely as close as is feasible to a randomized trial as is possible,” said Don C. Rockey, MD, a professor of medicine at the Medical University of South Carolina, Charleston, in an interview.
“From the clinician’s perspective, it is important to take away that there may be differences among the DOACs in terms of where in the GI tract the bleeding occurs,” said Dr. Rockey. In the study, the greatest differences appeared to be in the upper GI tract, with rivaroxaban outpacing apixaban and dabigatran. In patients who are at risk for upper GI bleeding, it may be reasonable to consider use of dabigatran or apixaban, he suggested.
“A limitation of the study is that it is likely underpowered overall,” said Dr. Rockey. It also wasn’t clear how many deaths occurred either directly from GI bleeding or as a complication of GI bleeding, he said.The study also didn’t differentiate major bleeding among DOACs specifically in the upper or lower GI tract, Dr. Rockey added.
Other studies yield similar results
Dr. Ingason and Dr. Björnsson said their work complements previous studies, and Neena S. Abraham, MD, MSc , who has conducted a similar investigation to the new study, agreed with that statement.
Data from the last 4 years overwhelmingly show that rivaroxaban is most likely to cause GI bleeding, said Dr. Abraham, professor of medicine and a consultant with Mayo Clinic’s division of gastroenterology and hepatology, in an interview.
A comparative safety study Dr. Abraham coauthored in 2017 of rivaroxaban, apixaban, and dabigatran in a much larger U.S. cohort of 372,380 patients revealed that rivaroxaban had the worst GI bleeding profile. Apixaban was 66% safer than rivaroxaban and 64% safer than dabigatran to prevent gastrointestinal bleeding.
“I believe our group was the first to conduct this study and show clinically significant differences in GI safety of the available direct oral anticoagulants,” she said. Other investigators have since published similar results, and the topic of the new study needs no further investigation, according to Dr. Abraham.
“It is time for physicians to choose a better choice when prescribing a direct oral anticoagulant to their atrial fibrillation patients, and that choice is not rivaroxaban,” she said.
The Icelandic Centre for Research and the Landspítali University Hospital Research Fund provided funds for this study. Dr. Ingason, Dr. Björnsson, Dr. Rockey, and Dr. Abraham reported no disclosures.