WASHINGTON – Early resumption of antiplatelet agents or anticoagulants after a major gastrointestinal bleeding event is clearly associated with an increased risk of rebleeding, but a decreased risk of death, results from an observational study show.
Furthermore, anticoagulant treatment “is associated with a higher risk of rebleeding and death compared with antiplatelet treatment after a previous GI event,” Dr. Angel Lanas said to an overflow crowd at the annual Digestive Disease Week.
In a separate case-control study, Dr. Lanas and his associates recently reported that the risk of GI bleeding was twofold higher for anticoagulants than for low-dose aspirin in patients hospitalized for GI bleeding (Clin. Gastroenterol. Hepatol. 2015 May;13:906-12.e2. [doi:10.1016/j.cgh.2014.11.007])
The current study examined adverse events in a cohort of 160 patients who developed a major gastrointestinal bleed (GIB) while using anticoagulants and/or antiplatelet therapy between March 2008 and July 2013. Long-term interruption or short-term resumption of these treatments has important clinical implications and differences in the intrinsic risks between antiplatelet or anticoagulant users after drug resumption are not well established, said Dr. Lanas of the University of Zaragoza (Spain).
Drug use information was prospectively collected during the GIB event, with data during the follow-up period obtained from two different Spanish databases.
Treatment during the index bleeding event was continued without interruption in 11 patients and interrupted in 149 patients (93%). Among those whose therapy was interrupted, 21 (14%) never resumed therapy and 128 (86%) resumed therapy (118 patients within 15 days and 10 patients after 15 days). The 86% treatment resumption rate is much higher than the 40%-66% rates reported in previous studies, indicating that Spanish physicians restarted treatment quite early, Dr. Lanas observed.
The mean age at baseline was 76.6 years, 61.3% of patients were men, and half had a Charlson index score > 4. Median follow-up was 21.5 months (range 1-63 months).
Ischemic events did not differ between patients who did or did not restart anticoagulants or antiplatelets (16.4% vs. 14.3%; P value = .806). However, rebleeding occurred in 32% of patients who resumed therapy versus none who did not (P = .002), but deaths were higher in those who did not restart therapy (38.1% vs. 12.5%; P = .003), Dr. Lanas said.
These differences remain significant in Kaplan-Meier survival curves for death (P = .021) and rebleeding (P = .004).
A comparison of early therapy resumption (≤ 15 days) vs. delayed (mean delay 62 days) or no resumption revealed similar results. Early resumption was associated with a higher rate of rebleeding (32.2% vs. 9.7%; P = .012), but a lower rate of death (11% vs. 35.5%; P = .001), with no difference in ischemic events (17% vs. 13%; P = .586), Dr. Lanas said.
Again, the differences remain significant in Kaplan-Meier survival curves for death (P = .011) and rebleeding (P = .013).
When the investigators looked at rebleeding according to drug use, patients receiving anticoagulants vs. antiplatelets had significantly higher rates of rebleeding (34.7% vs. 20.5%; P = .043), death (22.2% vs. 10.2%; P = .038), and any event (68.1% vs. 52.3%; P = .043).
After adjustment for gender, age, Charlson index, diabetes, and arterial hypertension, the risk of rebleeding was more than threefold higher for dual antiplatelet and anticoagulant users than for antiplatelet-alone users (odds ratio, 3.45; P = .025) and was twofold higher for anticoagulant vs. antiplatelet users (OR, 2.07; P = .045), Dr. Lanas said.
Finally, an analysis of the cause of bleeding suggests the cause of rebleeding may be different from the original event and that there is a shift toward the lower GI tract, he added.
The index bleeding event was caused largely by an upper GI peptic ulcer in 48% of all 160 patients, with 43.7% of events due to lower GI diverticulosis, vascular lesions, ischemic, or other lesions. In contrast, peptic ulcers accounted for only 7% of rebleeding events, while lower GI events accounted for 72%. Proton pump inhibition use was evenly distributed in upper and lower GI bleeding, although effective endoscopic treatment may have influenced upper GI bleeds, Dr. Lanas said.
“The importance of this is that we may have very good therapy tools for the upper GI, but still we have problems controlling the bleeding from the lower GI,” he added.
During a discussion of the study, an audience member asked how many days clinicians should wait to restart anticoagulants or antiplatelets.