Abstract
DA-9601 extracted from Artemisia asiatica contains a bioactive compound – eupatilin – that can protect against gastric mucosal damage through anti-inflammatory and anti-oxidative properties and is approved for treating acute and chronic gastritis in Korea, but their ability to protect gastrointestinal (GI) bleeding caused by nonsteroidal anti-inflammatory drugs (NSAIDs) is unclear. We aimed to compare the protective effects of DA-9601 to those of proton pump inhibitors (PPI) and rebamipide against upper and lower GI bleeding in patients with rheumatoid arthritis (RA) undergoing long-term NSAIDs therapy using the Korean Health Insurance Review and Assessment database. In this nationwide retrospective cohort study, we evaluated patients with RA who concurrently received NSAIDs for >3 months with DA-9601, PPI, or rebamipide between January 2015 and December 2017. The index date was the date of NSAIDs initiation, and all patients were followed up until December 2020 to detect upper and lower GI bleeding. In total, 24,258 patients with RA were eligible, and 5468 (22.5%), 4417 (18.2%), and 14,373 (59.3%) received DA-9601, PPI, or rebamipide, respectively, on the index date. During follow-up, upper and lower GI bleeding occurred in 508 (2.1%) and 402 (1.6%) patients with RA, respectively. The incidence rate of upper and lower GI bleeding was 615/100,000 and 485/100,000 person-years, respectively. Among patients with RA receiving DA-9601, PPI, or rebamipide, the frequencies of NSAIDs-induced upper GI bleeding were 0.5%, 0.4%, and 1.2%, respectively. The frequencies of NSAIDs-induced lower GI bleeding were 0.4%, 0.4%, and 0.9%, respectively. The incidence of NSAIDs-induced upper GI bleeding in patients with RA receiving DA-9601, PPI, and rebamipide was 601/100,000, 705/100,000, and 596/100,000 person-years, respectively, while the incidence of NSAIDs-induced lower GI bleeding in the same groups was 449/100,000, 608/100,000, and 465/100,000 person-years, respectively. In the multivariate Cox regression analysis, no significant difference was observed in lower and upper GI bleeding hazards between patients with RA using DA-9601, PPI, and rebamipide. Our results suggest that DA-9601 may exhibit protection against NSAIDs-induced GI bleeding that is comparable to those of PPI and rebamipide in patients with RA.
Keywords: eupatilin, gastrointestinal hemorrhage, nonsteroidal anti-inflammatory agents, proton pump inhibitors, rheumatoid arthritis
1. Introduction
Nonsteroidal anti-inflammatory drugs (NSAIDs) are widely used to manage acute and chronic pain and alleviate symptoms of chronic inflammatory arthritis, such as rheumatoid arthritis (RA), owing to their anti-inflammatory and analgesic properties via cyclooxygenase (COX) inhibition and decreased production of prostaglandins.[1] NSAIDs are categorized into nonselective NSAIDs (nsNSAIDs) and selective COX-2 inhibitors based on the relative degree of inhibiting both COX isoforms – COX-1 and COX-2. COX-1 is constitutively expressed and involved in maintaining homeostasis in the human body. Contrastingly, COX-2 is inducible by various stimuli and contributes to the inflammatory process. In addition to their beneficial effects, COX inhibition by NSAIDs can lead to various adverse effects, such as gastrointestinal (GI) tract injury, cardiovascular diseases, and renal impairment, with GI mucosal damage being the most common.[2] Not only COX-1 but also COX-2 contribute to maintaining mucosal integrity and facilitate tissue damage repair in the GI tract;[3,4] thus, using nsNSAIDs and selective COX-2 inhibitors can lead to GI toxicity. The association between NSAIDs use and upper GI adverse events such as peptic ulcers is well recognized, and serious upper GI complications such as GI bleeding and perforation can occur in approximately 1% of individuals who use NSAIDs.[5] Lower GI side effects have received less attention than upper GI injuries in the past; however, a gradual increase in the incidence of lower GI tract damage has been reported during treatment with NSAIDs.[5] Considering the recent rise in the use of NSAIDs,[2] preventive measures for GI complications, such as the co-prescription of gastroprotective agents (GPAs), have become increasingly important.[6]
Among the GPAs, proton pump inhibitors (PPI) have the highest level of evidence for preventing upper GI complications in patients receiving long-term NSAIDs therapy because of their ability to suppress gastric acid secretion.[6] However, unlike upper GI complications, lower GI damage caused by NSAIDs has various etiologies rather than increased gastric acid secretion; thus, a recent meta-analysis reported that PPIs are ineffective or exacerbate this condition.[7] In addition, long-term PPI use increases the risk of various disorders, such as chronic kidney disease (CKD), dementia, fractures, and infections,[8] underscoring the need for caution in the widespread prescription of this medication. Rebamipide, another GPA, can enhance prostaglandin generation in the gastric mucosa, promoting the healing of peptic ulcers. Moreover, a recent multicenter pilot study in Korea revealed its potential to protect against upper and lower GI complications induced by NSAIDs.[9] DA-9601 extracted from Artemisia asiatica contains a bioactive compound – eupatilin – that can protect against gastric mucosal damage through anti-inflammatory and anti-oxidative properties.[10] It is approved for treating acute and chronic gastritis in Korea. In a recent experimental study, DA-9601 ameliorated the inflammatory response in an aceclofenac-induced acute enteritis model,[11] suggesting its potential efficacy against NSAIDs-induced lower GI complications. DA-9601 is widely used as a GPA in Korea; nonetheless, its protective efficacy against NSAIDs-induced GI damage compared with PPI or rebamipide is not well understood. Thus, we aimed to compare the protective effect of DA-9601 against upper and lower GI bleeding to those of PPI and rebamipide in patients with RA undergoing long-term NSAIDs therapy using a nationwide claims database.
2. Methods
2.1. Data source
Approximately 97% of the Korean population must enroll in the Korean National Health Insurance Service (NHIS). This program offers reimbursement for a portion of the total medical expenses, ranging from 70% to 95%, based on the review of claims submitted by medical institutions and the final reimbursement decision made by the Korean Health Insurance Review and Assessment Service (HIRA).[12] Thus, the HIRA database contains claims data for inpatient and outpatient medical services of nearly 50 million Korean citizens, including demographics (age, sex, and residential areas), type of medical institution, prescribed medications, laboratory and imaging tests, procedures, surgeries, and diagnostic codes, but does not provide individual patient test results. In the HIRA database, the diagnoses of individual insurance beneficiaries are recorded according to the seventh edition of the Korean Classification of Diseases (KCD-7), a modified version of the tenth edition of the International Classification of Diseases.
2.2. Study designs and patients
We conducted a nationwide retrospective cohort study using the customized HIRA database, where the beneficiary information was anonymized. This study analyzed patients with RA aged ≥ 20 who received newly prescribed NSAIDs continuously for ≥ 3 months between January 2015 and December 2017 while concurrently taking either
DA-9601, PPI, or rebamipide. The index date was the date of the first NSAIDs prescription. Only oral NSAIDs were included in the analysis; parenteral NSAIDs were not considered. The “continuous use” of NSAIDs and GPAs (DA-9601, PPI, or rebamipide) for ≥3 months was defined as having a medication possession ratio of ≥0.8 for each respective drug for 3 months. Patients with RA were identified as having relevant diagnostic codes (M05 or M06 in KCD-7) with concomitant prescription of disease-modifying antirheumatic drugs (DMARDs) before the index date, as previously described.[13] To ensure the identification of NSAIDs initiators, patients with RA who had used NSAIDs within 6 months before the index date were excluded. In addition, the following criteria were used to exclude patients with RA: patients with a GI bleeding history or risk factors for GI bleeding, such as GI malignancy, inflammatory bowel diseases, and alcohol-related diseases; those receiving GPAs other than DA-9601, PPI, or rebamipide; and those receiving a combination of DA-9601, PPI, and rebamipide. Figure 1 illustrates the flowchart for identifying eligible study participants. The Research and Ethical Review Board of Pusan National University Hospital approved this study. This was a retrospective study using anonymized claim data; thus, the requirement for informed consent was waived (IRB no. 2003-002-088, approval date: March 9, 2020).
Figure 1.
Flow chart of this study.
2.3. Covariates and outcomes
The following covariates were collected from the HIRA database: age, sex, type of NSAIDs (nsNSAIDs or selective COX-2 inhibitors), index year, prescriber specialty (rheumatologists or non-rheumatologists), medical institution (tertiary, secondary, or primary/other), comorbidities, medications for RA treatment, and other medications. The nsNSAIDs included aceclofenac, cinnoxicam, dexibuprofen, diclofenac, etodolac, ibuprofen, ketorolac, lornoxicam, loxoprofen, mefenamic acid, meloxicam, nabumetone, naproxen, nimesulide, morniflumate, piroxicam, pranoprofen, proglumetacin, propionic acid, sulindac, talniflumate, tiaprofenic acid, and zaltoprofen. Selective COX-2 inhibitors included celecoxib and etoricoxib, which were available during the study in Korea. Comorbidities included peptic ulcers, gastroesophageal reflux disease (GERD), diabetes mellitus (DM), hypertension, CKD, liver diseases, coronary artery diseases, heart failure (HF), dyslipidemia, and chronic pulmonary obstructive disease. Medications for RA included glucocorticoids (GCs), conventional synthetic DMARDs (csDMARDs) such as methotrexate (MTX), sulfasalazine, hydroxychloroquine, and leflunomide, and biological DMARDs. Biological DMARDs included infliximab, adalimumab, etanercept, golimumab, abatacept, tocilizumab, and rituximab. Medications that increase the risk of GI bleeding, such as clopidogrel, cilostazol, ticlopidine, warfarin, rivaroxaban, dabigatran, and selective serotonin uptake inhibitors, were categorized as other medications.
The primary outcome of this study was the occurrence of upper and lower GI bleeding. Upper GI bleeding was determined via an endoscopic examination in conjunction with any of the following diagnostic codes: K226, K228, K250, K252, K254, K256, K260, K262, K264, K266, K270, K272, K274, K276, K280, K282, K284, K286, K290, K3181, and K920 in KCD-7.[14,15] Lower GI bleeding was defined using the following relevant diagnostic codes in KCD-7: K5711, K5713, K5731, K5733, K625, K5521, K921, and K922.[14,15] All study patients were followed up until December 2020 to detect GI bleeding (study events). Censoring was defined as cases in which the study patients had their NHIS discontinued, discontinued the NSAIDs or GPAs such as DA-9601, PPI, and rebamipide, or reached the end of the study period (December 2020). If there was a switch from nsNSAIDs to selective COX-2 inhibitors or vice versa, the NSAIDs treatment was considered continued rather than discontinued.
2.4. Statistical analyses
We performed all statistical analyses using SAS version 9.4 (SAS Institute, Cary, NC, USA) and considered a P value of <0.5 significant. Descriptive statistics were used to characterize demographic and clinical data. Continuous and categorical variables were compared using Student t test, Chi square test, or Fisher exact test, as appropriate. The primary goal of our analysis was to compare the protective effects of PPI, rebamipide, and DA-9601 against NSAIDs-induced upper and lower GI bleeding in patients with RA. Upper and lower GI bleeding-free survival rates were estimated using Kaplan–Meier curves and compared using the log-rank test. We calculated the hazard ratios (HR) with 95% confidence intervals (CI) of PPI, rebamipide, and DA-9601 for upper and lower GI bleeding using multivariate Cox proportional hazard regression models adjusted for variables that exhibited significant differences between the PPI, rebamipide, and DA-9601-treated groups at baseline and clinically relevant variables.
3. Results
In total, 24,258 patients with RA were eligible based on the inclusion and exclusion criteria (Fig. 1). The baseline demographic, clinical, and medication data of the study patients are presented in Table 1. The mean age was 53.1 years, and most patients were prescribed selective COX-2 inhibitors (75.6%), whereas 24.4% received prescriptions for nsNSAIDs. Approximately 60% of the patients with RA were concurrently taking GCs, and the most commonly used csDMARDs were hydroxychloroquine (54.7%), followed by MTX (47.3%), sulfasalazine (23.5%), and leflunomide (8.8%). The most common comorbidity was GERD (28.9%), followed by hypertension (28.4%), peptic ulcer (26.4%), liver disease (19%), dyslipidemia (16.5%), and DM (13.3%). Approximately half of the patients were prescribed NSAIDs by a rheumatologist at the index date. The proportions of patients with RA receiving DA-9601, PPI, and rebamipide were 22.5% (n = 5468), 18.2% (n = 4417), and 59.3% (n = 14,373), respectively (Table 1). A significant difference was observed in age, frequency of NSAIDs type, index year, prescriber specialty, medical institution, comorbidities except for CKD, and medications for RA among the DA-9601-, PPI-, and rebamipide-treated groups. In particular, patients with RA receiving DA-9601 had a lower frequency of nsNSAIDs prescription, prescription of NSAIDs by a rheumatologist, concomitant GCs prescription, peptic ulcers, and GERD than those receiving PPI.
Table 1.
Baseline characteristics of study patients.
| Total (n = 24,258) |
DA-9601 (n = 5468) |
PPI (n = 4417) |
Rebamipide (n = 14,373) |
P value | |
|---|---|---|---|---|---|
| Age, years, mean ± SD | 53.1 ± 13.4 | 53.8 ± 13.2 | 55 ± 13.2 | 52.3 ± 13.5 | 0.0289 |
| Male, n (%) | 5843 (24.1) | 1358 (24.8) | 1041 (23.6) | 3444 (24) | 0.294 |
| Type of NSAIDs | |||||
| Nonselective NSAIDs, n (%) | 5922 (24.4) | 1142 (20.9) | 1265 (28.6) | 3515 (24.5) | <0.001 |
| Selective COX-2 inhibitors, n (%) | 18,336 (75.6) | 4326 (79.1) | 3152 (71.4) | 10,858 (75.5) | |
| Index year | |||||
| 2010, n (%) | 11,749 (48.4) | 3092 (56.5) | 1927 (43.6) | 6730 (46.8) | <0.001 |
| 2011, n (%) | 5053 (20.8) | 1147 (21) | 866 (19.6) | 3040 (21.2) | |
| 2012, n (%) | 4503 (18.6) | 791 (14.5) | 904 (20.5) | 2808 (19.5) | |
| 2013, n (%) | 2953 (12.2) | 438 (8) | 720 (16.3) | 1795 (12.5) | |
| Prescriber specialty | |||||
| Rheumatologist, n (%) | 12,497 (51.5) | 2237 (40.9) | 2328 (52.7) | 7932 (55.2) | <0.001 |
| Non-rheumatologist, n (%) | 11,761 (48.5) | 3231 (50.9) | 2089 (47.3) | 6441 (44.8) | |
| Institution | |||||
| Tertiary hospital, n (%) | 10,140 (41.8) | 1800 (32.9) | 1783 (40.4) | 6557 (45.6) | <0.001 |
| Secondary hospital, n (%) | 5705 (23.5) | 1480 (27.1) | 1068 (24.2) | 3157 (22) | |
| Primary hospital/ other, n (%) | 8413 (34.7) | 2188 (40) | 1566 (35.4) | 4659 (32.4) | |
| Comorbidities | |||||
| Peptic ulcer | 6394 (26.4) | 1028 (18.8) | 1761 (39.9) | 3605 (25.1) | <0.001 |
| GERD, n (%) | 7016 (28.9) | 1015 (18.8) | 3293 (74.6) | 2708 (18.8) | <0.001 |
| DM, n (%) | 3217 (13.3) | 717 (13.1) | 693 (15.7) | 1807 (12.6) | <0.001 |
| HTN, n (%) | 6888 (28.4) | 1605 (29.4) | 1432 (32.4) | 3851 (26.8) | <0.001 |
| CKD, n (%) | 122 (0.5) | 34 (0.6) | 25 (0.6) | 63 (0.4) | 0.213 |
| Liver diseases, n (%) | 4611 (19) | 987 (18.1) | 921 (20.9) | 2703 (18.8) | 0.001 |
| CAD, n (%) | 220 (0.9) | 44 (0.8) | 54 (1.2) | 122 (0.8) | 0.048 |
| HF, n (%) | 129 (0.5) | 18 (0.3) | 30 (0.7) | 81 (0.6) | 0.042 |
| Dyslipidemia, n (%) | 3999 (16.5) | 984 (18) | 829 (18.8) | 2186 (15.2) | <0.001 |
| COPD, n (%) | 1705 (7) | 380 (6.9) | 369 (8.4) | 956 (6.7) | 0.001 |
| Medications for RA | |||||
| GCs, n (%) | 15,443 (63.7) | 3139 (57.4) | 3009 (68.1) | 9295 (64.7) | <0.001 |
| MTX, n (%) | 11,474 (47.3) | 2333 (42.7) | 2327 (52.7) | 6814 (47.4) | <0.001 |
| SSZ, n (%) | 5693 (23.5) | 1143 (20.9) | 927 (21) | 3623 (25.2) | <0.001 |
| HCQ, n (%) | 13,269 (54.7) | 2989 (54.7) | 2231 (50.5) | 8049 (56) | <0.001 |
| LEF, n (%) | 2124 (8.8) | 376 (6.9) | 590 (13.4) | 1158 (8.1) | <0.001 |
| Biological DMARDs, n (%) | 325 (1.3) | 25 (1.1) | 89 (2) | 178 (1.2) | <0.001 |
| Other medications | |||||
| Aspirin, n (%) | 29 (0.1) | 9 (0.2) | 5 (0.1) | 15 (0.1) | 0.542 |
| Clopidogrel, n (%) | 352 (1.5) | 79 (1.4) | 102 (2.3) | 171 (1.2) | <0.001 |
| Cilostazol, n (%) | 194 (0.8) | 58 (1.1) | 38 (0.9) | 98 (0.7) | 0.025 |
| Ticlopidine, n (%) | 23 (0.1) | 4 (0.1) | 5 (0.1) | 14 (0.1) | 0.803 |
| Warfarin, n (%) | 95 (0.4) | 16 (0.3) | 29 (0.7) | 50 (0.3) | 0.007 |
| Rivaroxaban, n (%) | 14 (0.1) | 5 (0.1) | 6 (0.1) | 3 (0) | 0.01 |
| Dabigatran, n (%) | 2 (0) | 0 (0) | 0 (0) | 2 (0) | 0.503 |
| SSRI, n (%) | 346 (1.4) | 71 (1.3) | 66 (1.5) | 209 (1.5) | 0.651 |
CAD = coronary artery disease, CKD = chronic kidney disease, COPD = chronic obstructive pulmonary disease, COX-2 = cyclooxygenase-2, DM = diabetes mellitus, DMARDs = disease-modifying antirheumatic drugs, GCs = glucocorticoids, GERD = gastroesophageal reflux disease, HCQ = hydroxychloroquine, HF = heart failure, HTN = hypertension, LEF = leflunomide, MTX = methotrexate, NSAIDs = nonsteroidal anti-inflammatory drugs, PPI = proton pump inhibitors, RA = rheumatoid arthritis, SD = standard deviation, SSRI = selective serotine reuptake inhibitors, SSZ = sulfasalazine.
During the study, 508 (2.1%) and 402 (1.7%) cases of upper and lower GI bleeding occurred in patients with RA after NSAIDs administration. The incidence of NSAIDs-induced upper and lower GI bleeding in patients with RA was 615/100,000 and 485/100,000 person-years, respectively. Among patients with RA receiving DA-9601, PPI, or rebamipide, the frequencies of NSAIDs-induced upper GI bleeding were 0.5% (n = 120), 0.4% (n = 97), and 1.2% (n = 291), respectively. The frequencies of NSAIDs-induced lower GI bleeding were 0.4% (n = 90), 0.4% (n = 84), and 0.9% (n = 228), respectively. The incidence of NSAIDs-induced upper GI bleeding in patients with RA receiving DA-9601, PPI, and rebamipide was 601/100,000, 705/100,000, and 596/100,000 person-years, respectively, while the incidence of NSAIDs-induced lower GI bleeding in the same groups was 449/100,000, 608/100,000, and 465/100,000 person-years, respectively. In the log-rank test, no significant difference existed in the time to upper (P = .489) and lower (P = .356) GI bleeding among the DA-9601-, PPI-, and rebamipide-treated groups.
The risk factors for upper and lower GI bleeding after NSAIDs use in patients with RA are detailed in Tables 2 and 3, respectively. After adjusting for confounding factors, the risk of NSAIDs-induced upper GI bleeding in patients with RA did not significantly differ among the DA-9601-, PPI-, and rebamipide-treated groups (Table 2). Older age (HR = 1.01, P = .001), underlying peptic ulcer (HR = 1.54, P < .001), GERD (HR = 1.41, P = .001), liver diseases (HR = 1.32, P = .009), concomitant use of GCs (HR = 1.38, P = .002), and cilostazol (HR = 1.96, P = .031) were independently associated with increased risk for NSAIDs-induced upper GI bleeding in patients with RA. In addition, multivariate Cox regression models revealed that the risk of NSAIDs-induced lower GI bleeding did not differ between the DA-9601-, PPI-, and rebamipide-treated groups (Table 3). Older age (HR = 1.17, P < .001), the male sex (HR = 1.56, P = .001), underlying DM (HR = 1.36, P = .021), HF (HR = 3.33, P = .001), and concomitant use of MTX (HR = 1.28, P = .02) were associated with a higher risk of lower GI bleeding. In contrast, NSAIDs prescribed by a rheumatologist had a significantly lower risk of lower GI bleeding (HR = 0.72, P = .007) after adjusting for confounding factors.
Table 2.
Risk factors for nonsteroidal anti-inflammatory drugs-induced upper gastrointestinal bleeding in patients with rheumatoid arthritis.
| Univariable | P value | Multivariable | P value | |
|---|---|---|---|---|
| HR (95% CI) | HR (95% CI) | |||
| Gastroprotective agents | ||||
| PPI | 1.11 (0.85–1.45) | 0.454 | 0.86 (0.64–1.15) | 0.302 |
| Rebamipide | 0.96 (0.78–1.19) | 0.727 | 1 (0.81–1.25) | 0.983 |
| DA-9601 | Ref. | Ref. | ||
| Age, years | 1.02 (1.02–1.03) | <0.001 | 1.01 (1.01–1.02) | 0.001 |
| Male | 1.06 (0.87–1.3) | 0.552 | 1.12 (0.91–1.38) | 0.276 |
| Type of NSAIDs | ||||
| Nonselective NSAIDs | 1.06 (0.87–1.3) | 0.545 | 1.01 (0.82–1.26) | 0.902 |
| Selective COX-2 inhibitors | Ref. | Ref. | ||
| Index year | ||||
| 2010 | 4.77 (2.61–8.72) | <0.001 | 5.07 (2.77–9.29) | <0.001 |
| 2011 | 3.07 (1.64–5.75) | 0.001 | 3.22 (1.72–6.03) | <0.001 |
| 2012 | 2.08 (1.08–4.01) | 0.029 | 2.11 (1.09–4.07) | 0.027 |
| 2013 | Ref. | Ref. | ||
| Prescriber specialty | ||||
| Rheumatologist | 0.88 (0.74–1.05) | 0.166 | 0.97 (0.78–1.2) | 0.762 |
| Non-rheumatologist | Ref. | Ref. | ||
| Institution | ||||
| Secondary hospital | 1.17 (1.04–1.32) | 0.01 | 0.95 (0.75–1.21) | 0.682 |
| Primary hospital/other | 1.25 (1.13–1.65) | <0.001 | 0.98 (0.77–1.24) | 0.867 |
| Tertiary hospital | Ref. | |||
| Comorbidities | ||||
| Peptic ulcer | 1.73 (1.45–2.07) | <0.001 | 1.54 (1.28–1.86) | <0.001 |
| GERD | 1.46 (1.22–1.75) | <0.001 | 1.41 (1.15–1.73) | 0.001 |
| DM | 1.44 (1.14–1.81) | 0.002 | 1.05 (0.82–1.34) | 0.707 |
| HTN | 1.68 (1.4–2) | <0.001 | 1.25 (1.02–1.53) | 0.034 |
| Liver diseases | 1.43 (1.17–1.75) | 0.001 | 1.32 (1.07–1.63) | 0.009 |
| CAD | 2.02 (1.05–3.91) | 0.036 | 1.36 (0.69–2.65) | 0.376 |
| HF | 1.2 (0.39–3.72) | 0.757 | 0.85 (0.27–2.67) | 0.779 |
| Dyslipidemia | 1.27 (1.02–1.58) | 0.032 | 0.99 (0.78–1.24) | 0.903 |
| COPD | 1.26 (0.92–1.71) | 0.149 | 0.97 (0.71–1.32) | 0.83 |
| Medications | ||||
| GCs | 1.36 (1.12–1.65) | 0.002 | 1.38 (1.13–1.69) | 0.002 |
| MTX | 0.97 (0.81–1.15) | 0.696 | 0.92 (0.76–1.1) | 0.682 |
| SSZ | 0.88 (0.71–1.09) | 0.226 | 1 (0.76–1.1) | 0.356 |
| HCQ | 1.12 (0.94–1.33) | 0.226 | 1.12 (0.92–1.35) | 0.256 |
| LEF | 1.49 (1.15–1.93) | 0.003 | 1.3 (0.98–1.71) | 0.067 |
| Biological DMARDs | 1.02 (0.48–2.15) | 0.958 | 1.04 (0.49–2.21) | 0.926 |
| Clopidogrel | 1.97 (1.16–3.35) | 0.013 | 1.22 (0.7–2.12) | 0.482 |
| Cilostazol | 2.77 (1.53–5.04) | 0.001 | 1.96 (1.06–3.63) | 0.031 |
CI = confidence intervals, COPD = chronic obstructive pulmonary disease, COX-2 = cyclooxygenase-2, DM = diabetes mellitus, DMARDs = disease-modifying antirheumatic drugs, GCs = glucocorticoids, GERD = gastroesophageal reflux disease, HCQ = hydroxychloroquine, HF = heart failure, HR = hazard ratio, HTN = hypertension, LEF = leflunomide, MTX = methotrexate, NSAIDs = nonsteroidal anti-inflammatory drugs, PPI = proton pump inhibitors, SSZ = sulfasalazine.
Table 3.
Risk factors for nonsteroidal anti-inflammatory drugs-induced lower gastrointestinal bleeding in patients with rheumatoid arthritis.
| Univariable | P value | Multivariable | P value | |
|---|---|---|---|---|
| HR (95% CI) | HR (95% CI) | |||
| Gastroprotective agents | ||||
| PPI | 1.28 (0.95–1.72) | 0.107 | 1.17 (0.84–1.63) | 0.345 |
| Rebamipide | 1.01 (0.79–1.29) | 0.955 | 1.08 (0.84–1.38) | 0.542 |
| DA-9601 | Ref. | Ref. | ||
| Age, years | 1.03 (1.02–1.03) | <0.001 | 1.02 (1.01–1.03) | <0.001 |
| Male | 1.53 (1.24–1.89) | <0.001 | 1.56 (1.26–1.93) | <0.001 |
| Type of NSAIDs | ||||
| Nonselective NSAIDs | 0.98 (0.78–1.24) | 0.883 | 0.91 (0.71–1.16) | 0.46 |
| Selective COX-2 inhibitors | Ref. | Ref. | ||
| Index year | ||||
| 2010 | 5.69 (2.68–12.08) | <0.001 | 6.03 (2.83–12.86) | <0.001 |
| 2011 | 4.95 (2.29–10.7) | <0.001 | 5.21 (2.41–11.28) | <0.001 |
| 2012 | 2.58 (1.15–5.81) | 0.022 | 2.65 (1.18–5.97) | 0.018 |
| 2013 | Ref. | Ref. | ||
| Prescriber specialty | ||||
| Rheumatologist | 0.79 (0.65–0.97) | 0.021 | 0.72 (0.57–0.91) | 0.007 |
| Non-rheumatologist | Ref. | Ref. | ||
| Institution | ||||
| Secondary hospital | 1.1 (0.86–1.41) | 0.458 | 0.93 (0.71–1.2) | 0.576 |
| Primary hospital/other | 0.91 (0.72–1.14) | 0.395 | 0.69 (0.53–0.9) | 0.006 |
| Tertiary hospital | Ref. | |||
| Comorbidities | ||||
| Peptic ulcer | 1.37 (1.12–1.69) | 0.003 | 1.23 (0.99–1.52) | 0.056 |
| GERD | 1.24 (1.01–1.53) | 0.042 | 1.12 (0.88–1.43) | 0.346 |
| DM | 1.89 (1.49–2.39) | <0.001 | 1.36 (1.05–1.77) | 0.021 |
| HTN | 1.54 (1.26–1.88) | <0.001 | 0.99 (0.79–1.26) | 0.994 |
| Liver diseases | 1.24 (0.98–1.57) | 0.074 | 1.07 (0.84–1.37) | 0.559 |
| CAD | 1.66 (0.74–3.71) | 0.219 | 0.9 (0.39–2.05) | 0.798 |
| HF | 4.84 (2.5–9.37) | <0.001 | 3.33 (1.67–6.56) | 0.001 |
| Dyslipidemia | 1.56 (1.24–1.97) | 0.001 | 1.25 (0.98–1.6) | 0.077 |
| COPD | 1.3 (0.92–1.83) | 0.137 | 1 (0.71–1.42) | 0.999 |
| Medications | ||||
| GCs | 1.07 (0.87–1.32) | 0.497 | 0.98 (0.79–1.23) | 0.88 |
| MTX | 1.22 (0.99–1.41) | 0.052 | 1.28 (1.04–1.58) | 0.02 |
| SSZ | 1.08 (0.86–1.35) | 0.52 | 1.18 (0.93–1.49) | 0.178 |
| HCQ | 1.03 (0.85–1.26) | 0.766 | 1.1 (0.9–1.36) | 0.355 |
| LEF | 1.37 (1.01–1.85) | 0.042 | 1.29 (0.93–1.77) | 0.125 |
| Biological DMARDs | 1.29 (0.61–2.72) | 0.506 | 1.34 (0.63–2.87) | 0.447 |
| Clopidogrel | 1.79 (0.96–3.36) | 0.068 | 1.02 (0.53–1.95) | 0.948 |
| Cilostazol | 2.51 (1.25–5.05) | 0.01 | 1.6 (0.78–3.27) | 0.197 |
CI = confidence intervals, COPD = chronic obstructive pulmonary disease, COX-2 = cyclooxygenase-2, DM = diabetes mellitus, DMARDs = disease-modifying antirheumatic drugs, GCs = glucocorticoids, GERD = gastroesophageal reflux disease, HCQ = hydroxychloroquine, HF = heart failure, HR = hazard ratio, HTN = hypertension, LEF = leflunomide, MTX = methotrexate, NSAIDs = nonsteroidal anti-inflammatory drugs, PPI = proton pump inhibitors, SSZ = sulfasalazine.
4. Discussion
To the best of our knowledge, this is the first study to compare the protective effects of various GPAs against NSAIDs-induced GI bleeding in patients with RA using a nationwide claims database. No clinically substantial difference was observed in long-term (≥3 months) NSAIDs-induced upper and lower GI bleeding among the DA-9601-, PPI-, and rebamipide-treated patients with RA. Thus, DA-9601 may have protective effects similar to those of PPI and rebamipide for NSAIDs-induced serious GI complications in patients with RA. We also observed that approximately 2% of patients with RA on long-term NSAIDs therapy experienced GI bleeding, underscoring the importance of monitoring NSAIDs-induced GI complications in RA management.
The key finding of this study is that DA-9601 may exhibit prevention similar to those of PPIs and rebamipide against NSAIDs-induced upper and lower GI bleeding. DA-9601 demonstrated a protective effect against NSAIDs-induced gastroduodenal injury that was not inferior to that of misoprostol in previous double-blinded randomized clinical trials[16,17] with a significant reduction in small bowel bleeding in patients taking aspirin in a previous study using the Korean claim database.[18] However, unlike our study, which focused on patients with RA, these previous studies did not exclusively analyze individuals with arthritis. Previous studies have suggested that arthritis such as RA or osteoarthritis (OA) may render the GI tract more susceptible to damage by NSAIDs;[19–21] therefore, our data provide more pertinent information for preventive measures against serious GI complications in patients with arthritis receiving long-term NSAIDs. Contrarily, Kim et al reported that rebamipide, misoprostol, PPIs, and histamine-2 receptor blockers (H2RB) were associated with a decreased risk of “occult” GI bleeding – a drop of ≥2 g/dL in hemoglobin level – in individuals with RA or OA who had been taking NSAIDs for at least 1 month; however, DA-9601 was unassociated.[22] A possible reason for this contradictory result could be the difference in endpoints between our study, which set “overt” GI bleeding as the study outcome, and the study by Kim et al.
Prescribing NSAIDs at the lowest effective dose and for the shortest possible duration is crucial to minimize GI injury;[6] however, their long-term use in clinical practice is common, particularly in patients with RA.[23,24] The mechanisms underlying NSAIDs-induced upper and lower GI mucosal damage differ, warranting the need for distinct prevention strategies for each complication. Gastric acid is a key factor in NSAIDs-induced gastroduodenal mucosal injury; thus, potent acid suppressants such as PPI are strongly recommended for patients receiving long-term NSAIDs therapy with a high risk of peptic ulcer disease (PUD).[6] NSAIDs-induced enteropathy is independent of gastric acid, and its pathogenesis includes multiple factors such as oxidative stress, enterohepatic circulation of NSAIDs, microbe-host interactions, and aberrant immune responses,[25] which are mechanisms suggesting that acid suppression is not sufficient to prevent this complication. PPIs are 1 of the most widely used prophylactic medications for NSAIDs-induced PUD; nonetheless, recent evidence suggests that PPI can induce bacterial overgrowth in the small bowel, subsequently increasing the risk of lower GI bleeding, which is more pronounced in patients using NSAIDs.[26] In addition, as mentioned above, long-term PPI use can be accompanied by various adverse effects in organs beyond the GI tract. Thus, achieving a risk-benefit balance when choosing PPIs for the long-term prevention of NSAIDs-induced GI complications is essential. Otherwise, DA-9601 is an herbal extract with a <2% side effect frequency in previous clinical trials,[27,28] indicating that it may be safer than PPI. Therefore, our results suggest that DA-9601 could be a favorable option for long-term prophylactic therapy for NSAIDs-induced serious upper and lower GI complications in patients with RA.
This study also estimated the incidence and risk factors of NSAID-induced upper and lower GI bleeding in patients with RA. The frequency of GI complications caused by NSAIDs in patients with RA and whether this frequently is higher than in patients with other joint disorders remain poorly characterized.[29] In addition, Watanabe reported that antitumor necrosis factor-α therapy could reduce the incidence of NSAIDs enteropathy in patients with RA;[30] however, comprehensive data analyzing the risk factors for NSAIDs-induced GI injury in these populations are still lacking. Therefore, our study provides valuable information regarding the epidemiology of NSAIDs-induced GI complications in RA. We revealed that the frequencies of NSAIDs-induced upper and lower GI bleeding were 2.1% and 1.7%, with incidences of 615/100,000 and 485/100,000 person-years, respectively, in patients with RA. Similar to our results, Laine et al reported that the incidence of serious lower GI events in patients with RA was 0.41 per 100 person-years in the rofecoxib-treated group and 0.89 per 100 person-years in the naproxen-treated group, representing a 0.9% annual rate of severe lower GI complications in patients receiving naproxen.[31] As previously established, old age, underlying peptic ulcers, GERD, and concomitant use of glucocorticoids and cilostazol were associated with a higher risk of NSAIDs-induced lower GI bleeding in patients with RA in our study. Notably, old age, the male sex, DM, HF, and the concomitant use of MTX were significant risk factors for NSAIDs-induced lower GI bleeding in patients with RA. Thus, these findings suggest the need to examine the occurrence of GI complications in patients with RA exhibiting these characteristics when using long-term NSAIDs therapy.
This study had several limitations that must be addressed. First, the HIRA database does not include individual clinical information, such as Helicobacter pylori infection and smoking history, which could be confounding factors for NSAIDs-induced GI bleeding. Second, NSAIDs-induced upper and lower GI bleeding was defined solely based on diagnostic codes in the claims database rather than the results of endoscopic findings, possibly leading to an underestimation of our study outcomes. This notion was also pointed out by Ingason et al, who reported that detecting GI bleeding through diagnostic codes is approximately 53% less sensitive than using endoscopy.[32] Third, our study did not include patients with RA receiving NSAIDs therapy without concurrently taking any GPA. Therefore, we could not estimate the net effect of DA-9601, PPI, and rebamipide on NSAIDs-induced GI bleeding. Fourth, our study was retrospective; therefore, it may not have captured all potential confounding variables that could have influenced NSAIDs-induced GI bleeding in patients with RA, and there may be limitations in establishing causality based on our findings.
5. Conclusion
This nationwide retrospective cohort study revealed no significant difference in the incidence of NSAIDs-induced upper and lower GI bleeding among patients with RA receiving DA-9601, PPI, and rebamipide who were on long-term NSAIDs therapy, suggesting that DA-9601 may offer a protective effect against NSAIDs-induced GI bleeding comparable to those of PPI and rebamipide. When long-term NSAIDs therapy is unavoidable, using GPAs with better long-term risk-benefit profiles to prevent serious GI complications in patients with RA is essential. Our findings provide valuable insights into the optimal prophylactic GPAs for NSAIDs-induced GI damage in patients with RA. Considering the limitations of this study, further large-scale prospective studies are required to validate our findings.
Author contributions
Formal analysis: Min Wook So.
Funding acquisition: Seung-Geun Lee.
Investigation: Min Wook So, Seung-Geun Lee.
Methodology: Min Wook So, Seung-Geun Lee.
Validation: Min Wook So, Seung-Geun Lee.
Writing – original draft: Min Wook So.
Data curation: Aran Kim.
Software: Aran Kim.
Writing – review & editing: Aran Kim, Seung-Geun Lee.
Conceptualization: Seung-Geun Lee.
Project administration: Seung-Geun Lee.
Resources: Seung-Geun Lee.
Supervision: Seung-Geun Lee.
Abbreviations:
- CI
- confidence intervals
- CKD
- chronic kidney disease
- COX
- cyclooxygenase
- csDMARDs
- conventional synthetic DMARDs
- DM
- diabetes mellitus
- DMARDs
- disease-modifying antirheumatic drugs
- GCs
- glucocorticoids
- GERD
- gastroesophageal reflux disease
- GI
- gastrointestinal
- GPAs
- gastroprotective agents
- H2RB
- histamine-2 receptor blockers
- HF
- heart failure
- HIRA
- Health Insurance Review and Assessment Service
- HR
- hazard ratios
- KCD-7
- the seventh edition of the Korean Classification of Diseases
- MTX
- methotrexate
- NHIS
- National Health Insurance Service
- NSAIDs
- nonsteroidal anti-inflammatory drugs
- nsNSAIDs
- non NSAIDs
- OA
- osteoarthritis
- PPI
- proton pump inhibitors
- PUD
- peptic ulcer disease
- RA
- rheumatoid arthritis
This study was supported in part by Dong-A pharmaceutical company.
This study was approved by the Research and Ethical Review Board of Pusan National University Hospital, which waived the requirement for informed consent because of the retrospective study design and anonymity of the extracted data (IRB no. 2003-002-088).
All authors declare no conflicts of interest.
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.
How to cite this article: So MW, Kim A, Lee S-G. DA-9601 has protective effects comparable to those of proton pump inhibitor and rebamipide against nonsteroidal anti-inflammatory drugs-induced upper and lower gastrointestinal bleeding in patients with rheumatoid arthritis: A nationwide study using Korean Health Insurance Review and Assessment Service database. Medicine 2024;103:27(e38801).
Contributor Information
Min Wook So, Email: thalsdnrso@naver.com.
Aran Kim, Email: solees17@naver.com.
References
- [1].Koo BS, Hwang S, Park SY, Shin JH, Kim TH. The relationship between long-term use of nonsteroidal anti-inflammatory drugs and kidney function in patients with ankylosing spondylitis. J Rheum Dis. 2023;30:126–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [2].Park SC, Chun HJ, Kang CD, Sul D. Prevention and management of non-steroidal anti-inflammatory drugs-induced small intestinal injury. World J Gastroenterol. 2011;17:4647–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [3].Sigthorsson G, Simpson RJ, Walley M, et al. COX-1 and 2, intestinal integrity, and pathogenesis of nonsteroidal anti-inflammatory drug enteropathy in mice. Gastroenterology. 2002;122:1913–23. [DOI] [PubMed] [Google Scholar]
- [4].Shin SJ, Noh CK, Lim SG, Lee KM, Lee KJ. Non-steroidal anti-inflammatory drug-induced enteropathy. Intest Res. 2017;15:446–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [5].Sostres C, Gargallo CJ, Lanas A. Nonsteroidal anti-inflammatory drugs and upper and lower gastrointestinal mucosal damage. Arthritis Res Ther. 2013;15(Suppl 3):S3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].Joo MK, Park CH, Kim JS, et al. Korean College of Helicobacter Upper Gastrointestinal Research. Clinical guidelines for drug-related peptic ulcer, 2020 Revised Edition. Gut Liver. 2020;14:707–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].Zhang X, Xiao X, Chen PR, Li YN, Lv XH, Yang JL. Proton pump inhibitors increase the risk of nonsteroidal anti-inflammatory drug-related small-bowel injury: a systematic review with meta-analysis. Clin Transl Gastroenterol. 2023;14:e00588. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [8].Gwee KA, Goh V, Lima G, Setia S. Coprescribing proton-pump inhibitors with nonsteroidal anti-inflammatory drugs: risks versus benefits. J Pain Res. 2018;11:361–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Oh DJ, Yoon H, Kim HS, et al. The effect of rebamipide on non-steroidal anti-inflammatory drug-induced gastro-enteropathy: a multi-center, randomized pilot study. Korean J Intern Med. 2022;37:1153–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [10].Park SW, Oh TY, Kim YS, et al. Artemisia asiatica extracts protect against ethanol-induced injury in gastric mucosa of rats. J Gastroenterol Hepatol. 2008;23:976–84. [DOI] [PubMed] [Google Scholar]
- [11].Kim JH, Shin CY, Jang SW, et al. Anti-inflammatory effects of DA-9601, an extract of Artemisia asiatica, on aceclofenac-induced acute enteritis. Korean J Physiol Pharmacol. 2021;25:439–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Kim HK, Song SO, Noh J, Jeong IK, Lee BW. Data configuration and publication trends for the Korean National health insurance and health insurance review & assessment database. Diabetes Metab J. 2020;44:671–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [13].Sung YK, Cho SK, Choi CB, Bae SC. Prevalence and incidence of rheumatoid arthritis in South Korea. Rheumatol Int. 2013;33:1525–32. [DOI] [PubMed] [Google Scholar]
- [14].Lin CC, Hu HY, Luo JC, et al. Risk factors of gastrointestinal bleeding in clopidogrel users: a nationwide population-based study. Aliment Pharmacol Ther. 2013;38:1119–28. [DOI] [PubMed] [Google Scholar]
- [15].Peng YL, Hu HY, Luo JC, Hou MC, Lin HC, Lee FY. Alendronate, a bisphosphonate, increased upper and lower gastrointestinal bleeding: risk factor analysis from a nationwide population-based study. Osteoporos Int. 2014;25:1617–23. [DOI] [PubMed] [Google Scholar]
- [16].Lee KN, Lee OY, Choi MG, et al. Prevention of NSAID-associated gastroduodenal injury in healthy volunteers-a randomized, double-blind, multicenter study comparing DA-9601 with misoprostol. J Korean Med Sci. 2011;26:1074–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [17].Lee OY, Kang DH, Lee DH, et al. A comparative study of DA-9601 and misoprostol for prevention of NSAID-associated gastroduodenal injury in patients undergoing chronic NSAID treatment. Arch Pharm Res. 2014;37:1308–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [18].Lee HS, Nam JH, Oh DJ, Ahn HJ, Lim YJ. Association between eupatilin and reduction in small bowel bleeding in aspirin users and aspirin plus acid suppressant users. Korean J Intern Med. 2023;38:484–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [19].Kato S, Takeuchi K. Alteration of gastric ulcerogenic and healing responses in rats with adjuvant-induced arthritis. Jpn J Pharmacol. 2002;89:1–6. [DOI] [PubMed] [Google Scholar]
- [20].McCafferty DM, Granger DN, Wallace JL. Indomethacin-induced gastric injury and leukocyte adherence in arthritic versus healthy rats. Gastroenterology. 1995;109:1173–80. [DOI] [PubMed] [Google Scholar]
- [21].Feng X, Tian M, Zhang W, Mei H. Gastrointestinal safety of etoricoxib in osteoarthritis and rheumatoid arthritis: a meta-analysis. PLoS One. 2018;13:e0190798. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [22].Kim TJ, Kim ER, Hong SN, et al. Effectiveness of acid suppressants and other mucoprotective agents in reducing the risk of occult gastrointestinal bleeding in nonsteroidal anti-inflammatory drug users. Sci Rep. 2019;9:11696. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [23].Grijalva CG, Chung CP, Stein CM, Mitchel EF, Jr., Griffin MR. Changing patterns of medication use in patients with rheumatoid arthritis in a Medicaid population. Rheumatology (Oxford). 2008;47:1061–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [24].Paglia MDG, Silva MT, Lopes LC, et al. Use of corticoids and non-steroidal anti-inflammatories in the treatment of rheumatoid arthritis: systematic review and network meta-analysis. PLoS One. 2021;16:e0248866. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [25].Zhang M, Xia F, Xia S, et al. NSAID-associated small intestinal injury: an overview from animal model development to pathogenesis, treatment, and prevention. Front Pharmacol. 2022;13:818877. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [26].Jung YS, Park JH, Park CH. Impact of proton pump inhibitors on the risk of small bowel or colorectal bleeding: a systematic review and meta-analysis. United Eur Gastroenterol J. 2023;11:861–73. In press. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [27].Seol SY, Kim MH, Ryu JS, Choi MG, Shin DW, Ahn BO. DA-9601 for erosive gastritis: results of a double-blind placebo-controlled phase III clinical trial. World J Gastroenterol. 2004;10:2379–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [28].Choi YJ, Lee DH, Choi MG, et al. Evaluation of the efficacy and safety of DA-9601 versus Its New Formulation, DA-5204, in patients with gastritis: phase III, randomized, double-blind, non-inferiority study. J Korean Med Sci. 2017;32:1807–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [29].Tacheci I, Bradna P, Douda T, et al. NSAID-induced enteropathy in rheumatoid arthritis patients with chronic occult gastrointestinal bleeding: a prospective capsule endoscopy study. Gastroenterol Res Pract. 2013;2013:268382. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [30].Watanabe T, Tanigawa T, Shiba M, et al. Anti-tumour necrosis factor agents reduce non-steroidal anti-inflammatory drug-induced small bowel injury in rheumatoid arthritis patients. Gut. 2014;63:409–14. [DOI] [PubMed] [Google Scholar]
- [31].Laine L, Connors LG, Reicin A, et al. Serious lower gastrointestinal clinical events with nonselective NSAID or coxib use. Gastroenterology. 2003;124:288–92. [DOI] [PubMed] [Google Scholar]
- [32].Ingason AB, Hreinsson JP, Agustsson AS, et al. Rivaroxaban is associated with higher rates of gastrointestinal bleeding than other direct oral anticoagulants a nationwide propensity score-weighted study. Ann Intern Med. 2021;174:1493–502. [DOI] [PubMed] [Google Scholar]