NSAIDs and Upper Gastrointestinal Bleeding

I appreciate the opportunity to respond to the letter from Dr. Küçükardalı and colleagues, which was prompted by the comments made in our review article regarding NSAID-related upper gastrointestinal bleeding (UGIB) in elderly patients [1]. Discussion concerning the outcome of gastrointestinal bleeding in this population is important in the light of recent data showing that admission rates for gastric and duodenal ulcer hemorrhage are increasing among older patients [2]. Several studies have documented that the risk of serious NSAID-associated gastrointestinal events increases with age. The reported magnitude of the risk associated with age ≥ 60 years has generally been in the range of 2.5 - 5.5 times that in patients younger than 60 [3] [4] [5], reaching as high as 12.7 for patients older than 75 years [6]. A recent observational study also found increased adjusted rate ratios for hospitalization due to upper gastrointestinal bleeding in elderly patients treated with nonselective NSAIDs and rofecoxib [7].

Küçükardalı et al. report on their comparison between NSAID-induced and other causes of upper gastrointestinal bleeding in an elderly cohort of patients (mean age approximately 70 years). After reviewing the records of patients admitted to the hospital over a period of 1 year, they selected 92 of 172 patients and found no differences in the outcomes between NSAID users and non-NSAID users with regard to transfusion requirements, length of hospital stay, and overall mortality. Several issues that may potentially influence the outcome of gastrointestinal bleeding are not dealt with in the report of their study and need to be discussed.

Firstly, it is unclear from the letter how the patients were selected for inclusion in the study, apart from having a diagnosis of upper gastrointestinal bleeding. Important aspects of study design, such as the criteria for inclusion (retrospective vs. prospective), the methodology of the selection process (consecutive, outpatient vs. in-patient gastrointestinal bleeding, etc.), and the criteria for excluding some 80 patients from the presentation of the results are not adequately explained in the letter.

Secondly, key information related to the type, dose, and duration of NSAID exposure is not provided in the letter. Although conflicting results have been reported on the relationship between gastrointestinal events and the duration of NSAID exposure [8], it is widely accepted that the risk of gastrointestinal events varies with the total dose and individual NSAIDs that patients are exposed to [9]. Similarly, and despite the ongoing debate surrounding the superiority in terms of gastroprotection of cyclooxygenase-2 (COX-2) inhibitors over conventional NSAIDs [10] [11], it would have been useful to know the differential rates of exposure to COX-2 inhibitors vs. traditional NSAIDs. A recent meta-analysis found a 39 % lower incidence of ulcer complications in patients taking COX-2 inhibitors in comparison with conventional NSAIDs [12].

Thirdly, there some information appears to be missing regarding patients exposed to NSAIDs in relation to the endoscopic diagnoses. In this group (n = 50), only 37 patients are accounted for in the description of bleeding lesions, and no information is provided about the source of bleeding in the other 13 patients.

Finally, two issues deserve mention in relation to the effect of confounding variables on the outcome of gastrointestinal bleeding in this cohort of patients. Although comorbid states did not significantly differ in the univariate analysis presented, the authors did not model the effects of all the potential independent factors through multiple regression analysis, so that the interaction between NSAID exposure, comorbidities, and outcome of gastrointestinal bleeding in these patients cannot be fully ascertained from their data (i. e., unadjusted mortality). Equally importantly, without the opportunity to control for many other important confounding factors such as smoking and alcohol consumption [13], use of endoscopic therapy [14], and drug exposures known to influence the occurrence and outcome of gastrointestinal bleeding (antisecretory medications, aspirin, anticoagulants, glucocorticoids, selective serotonin reuptake inhibitors, among others), the authors' conclusions should be viewed with skepticism. Prospective comparative studies controlling for the effect of independent variables on the outcome of gastrointestinal bleeding will provide the only reasonable evidence for the claim that there are significant differences in the outcome between NSAID-related and other causes of UGIB.

References

• 1 Rollhauser C, Fleischer D E. Nonvariceal upper gastrointestinal bleeding.  Endoscopy. 2002;  34 111-118
• 2 Higham J, Kang J Y, Majeed A. Recent trends in admissions and mortality due to peptic ulcer in England: increasing frequency of haemorrhage among older subjects.  Gut. 2002;  50 460-464
• 3 Gabriel S E, Jaakkimainen L, Bombardier C. Risk for serious gastrointestinal complications related to use of nonsteroidal anti-inflammatory drugs: a meta-analysis.  Ann Intern Med. 1991;  115 787-796
• 4 Silverstein F E, Graham D Y, Senior J R. et al . Misoprostol reduces serious gastrointestinal complications in patients with rheumatoid arthritis receiving nonsteroidal anti-inflammatory drugs: a randomized, double-blind, placebo-controlled trial.  Ann Intern Med. 1995;  123 241-249
• 5 Lanza L L, Walker A M, Bortnichak E A, Dreyer N A. Peptic ulcer and gastrointestinal hemorrhage associated with nonsteroidal anti-inflammatory drug use in patients younger than 65 years: a large health maintenance organization cohort study.  Arch Intern Med. 1995;  155 1371-1377
• 6 Hansen J M, Hallas J, Lauritsen J M, Bytzer P. Non-steroidal anti-inflammatory drugs and ulcer complications: a risk factor analysis for clinical decision-making.  Scand J Gastroenterol. 1996;  31 126-130
• 7 Mamdani M, Rochon P A, Juurlink D N. et al . Observational study of upper gastrointestinal haemorrhage in elderly patients given selective cyclo-oxygenase-2 inhibitors or conventional non-steroidal anti-inflammatory drugs.  Br Med J. 2002;  325 624-627
• 8 Laine L. Approaches to nonsteroidal anti-inflammatory drug use in the high-risk patient.  Gastroenterology. 2001;  120 594-606
• 9 Henry D, Lim L LY, Rodriguez L AG. et al . Variability in risk of gastrointestinal complications with individual non-steroidal anti-inflammatory drugs: results of a collaborative meta-analysis.  Br Med J. 1996;  312 1563-1566
• 10 Juni P, Rutjes A W, Dieppe P A. Are selective COX 2 inhibitors superior to traditional non steroidal anti-inflammatory drugs?.  Br Med J. 2002;  324 1287-1288
• 11 Budenholzer B R, Geis G S, Mamdani M. et al . Are selective COX 2 inhibitors superior to traditional NSAIDs?.  Br Med J. 2002;  325 161-163
• 12 Deeks J J, Smith L A, Bradley M D. Efficacy, tolerability, and upper gastrointestinal safety of celecoxib for treatment of osteoarthritis and rheumatoid arthritis: systematic review of randomised controlled trials.  Br Med J. 2002;  25 619-623
• 13 Andersen I B, Jorgensen T, Bonnevie O. et al . Smoking and alcohol intake as risk factors for bleeding and perforated peptic ulcers: a population-based cohort study.  Epidemiology. 2000;  11 434-439
• 14 Rollhauser C, Fleischer D E. Current status of endoscopic therapy for ulcer bleeding.  Baillière's Best Pract Res Clin Gastroenterol. 2000;  14 391-410