Current states of prevention of drug-induced gastroduodenal ulcer in clinical real practice

Background and Aim: Non-steroidal anti-inflammatory drugs (NSAIDs) or low-dose aspirin (LDA) are the most common causes of drug-induced gastroduodenal ulcer. We investigated preventive treatment with use of concomitant anti-ulcer drugs and the clinical features of gastroduodenal ulcer in cases treated with these drugs. Method: A total of 2,332 patients with gastroduodenal ulcer and 241 patients with bleeding gastroduodenal ulcer were classified into 3 groups: those taking LDA, those taking non-aspirin NSAIDs (hereinafter referred to as NSAIDs), and those taking neither aspirin nor NSAIDs. Chronological changes in the percentage of each group and the change over the past 15 years were investigated. The status of prevention of ulcer and clinical features were examined in 264 patients with gastroduodenal ulcer taking NSAIDs or LDA, including 107 bleeding cases, in the past 8 years. Results: From January 2002 to December 2017, the percentage of all patients taking LDA increased until 2013, but from 2014, the percentage of patients taking LDA decreased in those with a bleeding ulcer. The percentage of patients taking NSAIDs decreased from 2002 in those with a bleeding ulcer. Among the 264 patients with gastroduodenal ulcer and the 107 patients with a bleeding ulcer taking NSAIDs and LDA, 16 (6%) and 9 (8%), respectively, were receiving preventive treatment with concomitant anti-ulcer drugs. The percentages of patients taking LDA and other antiplatelet drugs in patients with bleeding gastroduodenal ulcer were significantly higher than those in patients with non-bleeding. Conclusion: Although the percentages of patients with gastroduodenal ulcer taking NSAIDs or LDA have not recently increased in real-world practice, preventive treatment in these patients is still low. This low rate of prevention in cases of non-bleeding and bleeding gastroduodenal ulcer suggests the need to enlighten physicians about preventive because drug of ulcer bleeding, and found that this was not inferior to stopping aspirin 18 . The results showed that patients who continued aspirin had lower all-cause mortality and lower mortality attributable to cardiovascular, cerebrovascular, or gastrointestinal complications, compared to patients who stopped aspirin 18 . Since stopping LDA after we a significant of use single in cases than in bleeding and we also a significant higher rate of use of LDA or other antiplatelet drugs in bleeding cases. These results may indicate that, in contrast to patients taking LDA or other antiplatelet drugs, those taking a single NSAID have a lower risk of gastroduodenal bleeding while under preventive treatment. These results also suggest that greater attention to gastroduodenal bleeding is needed in patients taking LDA or other antiplatelet drugs with concomitant preventive therapy.

Non-steroidal anti-inflammatory drugs (NSAIDs) are commonly used for pain in rheumatoid arthritis or orthopedic disease and are the most common cause of drug-induced gastroduodenal ulcer 1,2 . LDA is now commonly used for prevention of cardiovascular and cerebrovascular events with the advent of aging of society [3][4][5] . Worldwide trials have shown that LDA at 75-325 mg/day or other antiplatelet regimens offers beneficial protection against myocardial infarction, stroke, and death 4 . In contrast, a very low dose of aspirin (10 mg daily) decreases gastric mucosal prostaglandin levels and causes significant gastric mucosal damage 6 , and this may have increased the incidence of LDA-induced gastrointestinal mucosal injury [7][8][9][10] .
In cases of drug-induced gastroduodenal ulcer bleeding, drug withdrawal including LDA involves a high risk of cardiovascular and cerebrovascular events. Therefore, prevention of gastroduodenal ulcer bleeding in patients taking drugs including NSAIDs and LDA is clinically important 11 12 , but it is unclear to what extent preventive treatment is being used in such cases in real world practice.
Therefore, in the present study, we investigated chronological changes in use of NSAIDs and LDA in patients with gastroduodenal ulcer taking NSAIDs and LDA and examined the current status of prevention for these cases in clinical practice.

Statistical analysis
Data are expressed as mean ± SD. Categorical variables were compared by chi-square test, and continuous variables by Mann-Whitney test, with P < 0.05 considered to be significant.

Results
Chronological changes in patients with gastroduodenal ulcer taking NSAIDs and LDA For the patients with all or bleeding gastroduodenal ulcer, the mean ages of those taking LDA and NSAIDs were higher than those taking neither LDA nor NSAIDs (p<0.05). In these patients, the male/female ratio in those taking NSAIDs was lower than those in the other groups (p<0.05) ( Table 1) Preventive treatment in patients with gastroduodenal ulcer taking NSAIDs and LDA Among the 264 patients with gastroduodenal ulcer taking NSAIDs and LDA, 16 (6%) were receiving preventive treatment with concomitant anti-ulcer drugs, and 248 (94%) were not receiving this treatment. In the 107 patients with a bleeding ulcer taking NSAIDs and LDA, 9 (8%) were receiving preventive treatment, and 98 (92%) were not ( Table 2). In the 16 cases receiving preventive treatment, 11 (69%) and 5 (31%) were taking a half dose and a full dose of a proton pump inhibitors (PPIs) ( Table 3). In patients with non-bleeding and bleeding gastroduodenal ulcer, 85.7% and 22.2% were taking a single NSAID, and 14.3% and 77.8% were taking LDA or another antiplatelet drug , respectively. The ratio of bleeding gastroduodenal ulcer taking LDA or another antiplatelet drug were higher than that of non-bleeding cases, significantly (Table 4).

Discussion
The efficacy of LDA for prevention of cardiovascular and cerebrovascular diseases has been established 3,4,5 , but the risks of peptic ulcer complications increase in association with LDA use 6,[7][8][9][10] . A meta-analysis of 24 randomized controlled trials revealed that gastrointestinal hemorrhage occurred in 2.47% of patients taking aspirin compared with 1.42% taking placebo 7 .
We investigated the chronological changes in use of NSAIDs and LDA in patients with gastroduodenal ulcer since 2002. The percentage of these patients taking NSAIDs or LDA initially increased, but has decreased in recent years. High doses of histamine H2 receptor antagonist ( H2RAs), PPIs or prostaglandin analogs are recommended for prevention of NSAID or LDA-induced gastroduodenal ulcer, and PPIs are especially used widely as first-line drugs 11,12 . The spread of preventive use of PPIs may have caused the chronological changes in use of LDA and NSAIDs found in the current study.
However, in real world practice in our hospital, only 6% of patients with gastroduodenal ulcer taking NSAIDs and LDA were receiving preventive treatment with concomitant anti-ulcer drugs over the last 8 years, which indicates that use of this treatment is still insufficient.
Prevention of ulcer using PPIs is recommended, but some previous reports have suggested that this treatment is insufficient 13,14 . For example, it has been reported that 13% of cases had recurrence of gastroduodenal ulcer after treatment with 15 mg lansoprazole 15 . In cases of LDA-induced gastroduodenal ulcer and bleeding, discontinuation of LDA can increase the risk of cardiovascular and cerebrovascular diseases [16][17][18] , and continuation of LDA is recommended 12 . Sung et al. investigated continuation of aspirin therapy with PPIs after endoscopic control of ulcer bleeding, and found that this was not inferior to stopping aspirin 18 . The results showed that patients who continued aspirin had lower all-cause mortality and lower mortality attributable to cardiovascular, cerebrovascular, or gastrointestinal complications, compared to patients who stopped aspirin 18 . Since stopping LDA after gastroduodenal bleeding increases the risk of cardiovascular and cerebrovascular diseases, more thorough preventive treatment is needed. Recently, the efficacy of vonoprazan for prevention of NSAID and LDA-induced gastroduodenal ulcer has been reported in Japan 19,20 .
In cases under treatment with concomitant anti-ulcer drugs, we found that a significant higher rate of use of single NSAIDs in non-bleeding cases than in bleeding cases, and we also found a significant higher rate of use of LDA or other antiplatelet drugs in bleeding cases. These results may indicate that, in contrast to patients taking LDA or other antiplatelet drugs, those taking a single NSAID have a lower risk of gastroduodenal bleeding while under preventive treatment. These results also suggest that greater attention to gastroduodenal bleeding is needed in patients taking LDA or other antiplatelet drugs with concomitant preventive therapy.
In conclusion, although preventive treatment with PPIs is recommended, we found that this treatment is insufficiently applied in real world practice, and that few patients with gastroduodenal ulcer are receiving concomitant anti-ulcer drugs. However, stopping LDA after gastroduodenal bleeding increases the risk of cardiovascular and cerebrovascular diseases, and therefore, more thorough preventive treatment is needed in these cases. Tables   Due to technical  Chronological changes in use of NSAIDs and LDA in patients with bleeding due to gastroduodenal ulcer.

Supplementary Files
This is a list of supplementary files associated with this preprint. Click to download.