Warfarin Associated With Higher Upper GI Bleeding Rates Compared With DOACs

Carolyn Crist

August 26, 2022

Warfarin is associated with higher rates of upper gastrointestinal (GI) bleeding but not overall or lower gastrointestinal bleeding rates compared with direct oral anticoagulants (DOACs), according to a new nationwide report from Iceland.

In addition, warfarin is associated with higher rates of major gastrointestinal bleeding compared with apixaban.

Dr Arnar Ingason

"Although there has been a myriad of studies comparing GI bleeding rates between warfarin and DOACs, very few studies have compared upper and lower GI bleeding rates specifically," Arnar Ingason, MD, PhD, a gastroenterology resident at the University of Iceland and Landspitali University Hospital in Reykjavik, Iceland, told Medscape Medical News.

"Knowing whether the risk of upper and lower GI bleeding differs between warfarin and DOACs is important, as it can help guide oral anticoagulant selection," he said.

"Given that warfarin was associated with higher rates of upper GI bleeding compared to DOACs in our study, warfarin may not be optimal for patients with high risk of upper GI bleeding, such as patients with previous history of upper GI bleeding," Ingason added.

The study was published online in Clinical Gastroenterology and Hepatology.

Analyzing Bleed Rates

Ingason and colleagues analyzed data from electronic medical records for more than 7000 patients in Iceland who began a prescription for oral anticoagulants between 2014 and 2019. They used inverse probability weighting to yield balanced study groups and calculate the rates of overall, major, upper, and lower gastrointestinal bleeding. All events of gastrointestinal bleeding were manually confirmed by chart review.

Clinically relevant gastrointestinal bleeding was defined as bleeding that led to medical intervention, unscheduled physician contact, or temporary cessation of treatment. Upper gastrointestinal bleeding was defined as hematemesis or a confirmed upper gastrointestinal bleed site on endoscopy, whereas lower gastrointestinal bleeding was defined as hematochezia or a confirmed lower gastrointestinal bleed site on endoscopy. Patients with melena and uncertain bleeding site on endoscopy were classified as having a gastrointestinal bleed of unknown location.

Major bleeding was defined as a drop in hemoglobin of ≥ 20 g/L, transfusion of two or more packs of red blood cells, or bleeding into a closed compartment such as the retroperitoneum.

In total, 295 gastrointestinal bleed events were identified, with 150 events (51%) classified as lower, 105 events (36%) classified as upper, and 40 events (14%) of an unknown location. About 71% required hospitalization, and 63% met the criteria for major bleeding. Five patients died, including three taking warfarin and the other two taking apixaban and rivaroxaban.

Overall, warfarin was associated with double the rate of upper gastrointestinal bleeding, with 1.7 events per 100 person-years compared with 0.8 events per 100 person-years for DOACs. The rates of lower gastrointestinal bleeding were similar for the drugs.

Specifically, warfarin was associated with nearly 5.5 times higher rates of upper gastrointestinal bleeding compared with dabigatran (Pradaxa, Boehringer Ingelheim), 2.6 times higher than apixaban (Eliquis, Bristol Myers Squibb), and 1.7 times higher than rivaroxaban (Xarelto, Janssen). The risk for upper gastrointestinal bleeding also was higher in men taking warfarin.

Warfarin was associated with higher rates of major bleeding compared with apixaban, with 2.3 events per 100 person-years versus 1.5 events per 100 person-years. Otherwise, overall and major bleed rates were similar for users of warfarin and DOACs.

Dr Neena Abraham

"GI bleeding among cardiac patients on anticoagulants and antiplatelets is the fastest growing group of GI bleeders," Neena Abraham, MD, professor of medicine and a gastroenterologist at the Mayo Clinic in Scottsdale, Arizona, told Medscape Medical News.

Abraham, who wasn't involved with this study, runs a dedicated cardiogastroenterology practice and has studied these patients' bleeding risk for 20 years.

"This is a group that is ever increasing with aging baby boomers," she said. "It is anticipated by 2040 that more than 40% of the US adult population will have one or more cardiovascular conditions requiring the chronic prescription of anticoagulant or antiplatelet drugs."

Considering Future Research

In this study, peptic ulcer disease was a proportionally less common cause of upper gastrointestinal bleeding for warfarin, at 18%, compared with DOACs, at 39%. At the same time, the absolute propensity-weighted incidence rates of peptic ulcer-induced bleeding were similar, with 0.3 events per 100 person-years for both groups.

"As warfarin is not thought to induce peptic ulcer disease but rather promote bleeding from pre-existing lesions, one explanation may be that peptic ulcer disease almost always leads to overt bleeding in anticoagulated patients, while other lesions, such as mucosal erosions and angiodysplasias, may be more likely to lead to overt bleeding in warfarin patients due to a potentially more intense anticoagulation," Ingason said.

Ingason and colleagues now plan to compare GI bleeding severity between warfarin and DOACs. Previous studies have suggested that GI bleeding may be more severe in patients receiving warfarin than in those receiving DOACs, he said.

In addition, large studies with manual verification of gastrointestinal bleed events could better estimate the potential differences in the sources of upper and lower bleeding between warfarin and DOACs, Ingason noted.

"Some DOACs, specifically dabigatran, are known to have a mucosal effect on the luminal GI tract, as well as a systemic effect," Abraham said. "This pharmacologic effect may contribute to an increase in lower gastrointestinal bleeding in the setting of colonic diverticulosis or mucosal injuries from inflammatory processes."

Ongoing research should also look at different ways to reduce anticoagulant-related gastrointestinal bleeding among cardiac patients, she noted.

"Our research group continues to study the risk of cardiac and bleeding adverse events in patients prescribed to DOACs compared to those patients who receive a left atrial appendage occlusion device," Abraham said. "This device often permits patients at high risk of GI bleeding to transition off anticoagulant and antiplatelet drugs."

Clin Gastroenterol Hepatol. Published online August 13, 2022. doi:10.1016/j.cgh.2022.06.033. Abstract

The study was funded by the Icelandic Centre for Research and the Landspitali University Hospital Research Fund. The funders had no role in the design, conduct, or reporting of the study. The authors declared no competing interests. Abraham reported no relevant financial relationships.

Carolyn Crist is a health and medical journalist who reports on the latest studies for Medscape, MDedge, and WebMD.

For more news, follow Medscape on Facebook, Twitter, Instagram, YouTube, and LinkedIn.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.