Mucosal Erosions in Longterm Non-Steroidal Anti
Total Page:16
File Type:pdf, Size:1020Kb
334 Gut 1995; 36: 334-336 Mucosal erosions in longterm non-steroidal anti-inflammatory drug users: predisposition to ulceration and relation to Helicobacterpylori Gut: first published as 10.1136/gut.36.3.334 on 1 March 1995. Downloaded from A S Taha, R D Sturrock, R I Russell Abstract TABLE II Characteristics ofarthritis and details of The importance of erosions in longterm anti-rheumatoid second line drugs non-steroidal anti-inflammatory drug Erosions present Erosions absent (NSAID) users, their relevance to ulcera- (n=23) (n=27) tion and their relation to Helicobacter Osteoarthritis 3 5 pylori are unclear. This study assessed the Rheumatoid arthritis 20 22 Duration of arthritis (y) incidence of peptic ulcers in the presence median (interquartile range) 5 (3-11) 7 (3-11) or absence of erosions or H pylori in a Second line drugs Sulphasalazine 8 10 group of longterm NSAID users (n=50), Gold (intramuscular) 2 2 undergoing endoscopy at 0, 4, 12, and 24 Penicillamine 0 1 weeks while continuing with NSAIDs. Hydroxychloroquine 3 2 Ulcers diagnosed at baseline endoscopy were excluded. Ulcers developed in nine of 23 patients (390/o) with pre-existing erosions in longterm users is not clear. Also, erosions compared with six of 27 (22%) their link to ulceration has not been proved, without erosions (p<0.05). The group and their relation to Helicobacter pylori has not infected with H pylon (n=30) had a been defined. total of 18 patients (60%) with erosions, a We, therefore, aimed at investigating total of 12 ulcers (40/o), and eight ulcers the development of peptic ulcers in a group (27%) complicating previous erosions, of patients with chronic arthritis treated compared with five (25%, p<0-01), three with NSAIDs in the presence or absence of (15%, p<0.05), and one (5%/o, p<001) erosions, and to assess the relevance of respectively in patients not infected with H pylori to both erosions and ulcers in http://gut.bmj.com/ H pylori (n=20). Ulcer development was these patients. not influenced by the initial number of erosions but strongly associated with H pylori positive duodenal erosions. It is Patients and methods concluded that ulcers are more likely to Patients with osteoarthritis or rheumatoid develop in longterm NSAID users who arthritis, and aged 18 years or over, were have mucosal erosions or H pylori, or recruited from the Rheumatology Outpatient on September 28, 2021 by guest. Protected copyright. both. Clinic. NSAIDs had to be taken for at (Gut 1995; 36: 334-336) least four weeks before baseline endoscopy and to be continued throughout the study. Keywords: mucosal erosion, non-steroidal anti- Corticosteroids, cytotoxic drugs, or antiulcer inflammatory drug(s), Helicobacterpylori. treatment were not permitted. Endoscopy was performed after an overnight fast and intravenous midazolam was used for sedation. Gastric and duodenal erosions are the com- Patients with ulcers at baseline endoscopy monest endoscopic abnormalities related were excluded and those without ulcers were to acute exposure to non-steroidal anti-inflam- re-endoscoped at 4, 12, and 24 weeks. The matory drugs (NSAIDs).1 As they are primary end point was ulcer development at commonly repaired through processes of any stage of the study or the completion restitution and adaptation,2 erosions are often transient, and this in turn leads to the impres- TABLE III Details ofNSAID intake sion that they are comparatively trivial lesions.3 Erosions present Erosions absent Unlike the situation in subjects exposed to (n=23) (n=27) Departments of short courses of NSAIDs, the significance of Gastroenterology Type A S Taha Diclofenac 6 3 R I Russell Naproxen 5 6 TABLE I Demographic details of the study groups Indomethacin 4 7 and Rheumatology, Ketoprofen 2 2 Erosions absent Fenbufen 2 0 Royal Infirmary, Erosions present Nabumetone 1 3 Glasgow Number 23 27 Flurbiprofen 0 2 R D Sturrock Men 5 6 Others* 3 4 Women 18 21 Dosett 3 (2-3) 3 (2-3) Correspondence to: Age (y) median (interquartile Duration (years)4 2 (1-4) 2 (1-5) Dr A S Taha, GI Centre, range) 53 (43-59) 53 (43-61) Southem General Hospital, Smokers 10 11 *=One patient taking one of the following: ibuprofen, Glasgow G5 1 4TF. Alcohol drinkers 13 16 azapropazone, acematacin, or tiaprofenic acid. t=Dosage Accepted for publication Upper abdominal complaints 10 9 scale: 1, minimal dose; 2, submaximal; 3: maximal dose 24 June 1994 permitted. t=Median (interquartile range). Mucosal erosions in lonigterni nioni-steroidal anti-inflanmmatory dnig uisers: predisposition to ulceration and relation to Helicobacter pylori 335 TABLE IV Numbers ofpatients with erosionls or ulcers, in the presence (or absenice) of drugs (Table II), or in the types, doses, or H pylori at each endoscopy length of NSAID intake (Table III). Baseline 4 Weeks 12 Weeks 24 Weeks Table IV shows the numbers ofpatients with erosions or ulcers at each endoscopic assess- All erosions* 12 11 7 6 Gut: first published as 10.1136/gut.36.3.334 on 1 March 1995. Downloaded from All ulcers 0 8 5 2 ment. The study protocol excluded patients Ulcers plus erosions 0 2 3 1 with ulcers at baseline endoscopy. Erosions Erosions H pylon present (absent) 8 (4) 10 (1) 6 (1) 5 (1) were identified in 12 patients at the initial Ulcers endoscopy, and developed in 11 others on H pslon present (absent) 0 6 (2) 4 (1) 2 (0) subsequent examinations. They were recurrent * =Erosions developed in a total of 23 patients throughout the whole study, some of whom had in five patients, and disappeared altogether in lesions at more than one endoscopy (see text). eight patients. At each assessment, lesions were more likely to be found in patients of the required series of endoscopic examina- infected with H pylon. tions. An ulcer was defined as a mucosal Figure 1 shows ulcers developing in patients defect that is equal to or greater than 3 mm with or without previous erosions. Fifteen in diameter,4 5 and smaller lesions were con- ulcers were diagnosed in the study group as a sidered as erosions. Ulcers were described as whole (15 of 50, 30%0), with a median size of 4 deep (more than 3 mm) or comparatively mm (range 3-10 mm). Only five ulcers were superficial (2-3 mm in depth). Ulcer dimen- comparatively superficial and most (n= 10) sions were measured using the standard were deep. Ulcers affected the gastric body Olympus upper gastrointestinal endoscopy (n= 1), duodenum (n=5), gastric antrum biopsy forceps, with the fully open instrument (n=8), and both the duodenum and antrum being equivalent to 5 mm. The anatomical (n= 1). Most ulcers (nine of 15, 60%) locations of lesions were described as precisely developed in patients with previous or current as possible, their photographic appearances erosions. The median number of erosions (per recorded, and their distance from the teeth patient) appearing before an ulcer was measured. H pylorz was identified in gastric three (range 1-6) compared with four (1-12) antral biopsy specimens by both culture and erosions not followed by ulceration. Direct histological examination. The endoscopist was progression from an erosion to an ulcer was not aware of H pylon' status or the previous thought to have taken place in six of nine ulcer endoscopic findings, and the study patients patients with previous erosions. were re-endoscoped among a larger pool of Figure 2 shows ulcers and erosions classified arthritic patients to facilitate blinding and according to the presence or absence of further minimise familiarity between the H pylori. Patients with H pylori infection endoscopist and patients being repeatedly (n= 30) had significantly more erosions, ulcers, http://gut.bmj.com/ endoscoped. Statistical analyses included the and ulcers complicating previous erosions than X2 and Fisher's exact test where appropriate. patients not infected with Hpylori (n=20). It is The study was approved by the local ethics worth noting that of six patients with duodenal committee. erosions found in the study group as a whole five were positive for H pylon; all these five patients (100%) developed ulceration at Results subsequent endoscopic examinations (two on September 28, 2021 by guest. Protected copyright. A total of 50 patients were recruited. They gastric and three duodenal ulcers). This is were divided into two groups according to compared with only three ulcers developing in whether they had or had not had erosions at 13 patients (23%) with Hpylori positive gastric any stage of the study. Table I shows their erosions (p<0001) Also gastric submucosal demographic details. Patients in both groups haemorrhages were found in a total of five were comparable with respect to their ages, patients: three were infected with Hpylonr, and sex, smoking, and drinking habits. Also, there two of these three subsequently developed were no significant differences in the character- gastrlc eroslons. istics of patients' arthritis, use of second line Discussion 40 This study shows that longterm users of 35 NSAIDs with mucosal erosions are more likely to develop ulceration than patients without 30 these lesions. Also, the majority of both 0 25 - erosions and ulcers in longterm NSAID users 20 develop in patients infected with H pylon'. The (DI_.a) initial number of erosions in the individual 15 patient does not seem to influence the risk of 10 ulceration. In the presence of H pylori, 5 however, duodenal erosions have a stronger association with subsequent ulceration than All patients Patients with Patients without their gastric counterparts. (n=50) previous previous erosions erosions A variety of types of injury develop after the (n=23) (n=27) administration of NSAIDs, ranging from Figure 1: Ulcer development (%w) in patients with or petechial haemorrhages to erosions, and less without erosions.