Sucralfate treatment of nonsteroidal anti-inflammatory drug-induced gastrointestinal symptoms and mucosal damage

Sucralfate Treatment of Nonsteroidal Anti-Inflammatory Drug-Induced Gastrointestinal Symptoms and Mucosal Damage

JACQUES R. CALDWELL, M.D.

Gainesville, Florida

SANFORD H. ROTH, M.D.

Phoenix, Arizona

WALLACE C. WU, M.B., B.S. ELLIOTT L. SEMBLE, M.D. DONALD 0. CASTELL, M.D.

Winston-Salem, Notth Carolina

MATTHEW D. HELLER, M.D.

Lynnfield, Massachusetts

WILLIAM H. MARSH, M.D.

Charleston, South Carolina

From the University of Florida, Gainesville, Florida; Saint Luke’s Hospital Medical Center, Phoenix, Ari- zona; Bowman Gray School of Medicine, Wake Forest University, Winston-Salem, North Carolina; Arthritis Associates, Inc., Lynnfieid, Massachu- setts; and the Medical University of South Carolina, Charleston, South Carolina. This study was sup- ported by funds from Marion Laboratories, Inc., Kansas City, Missouri. Requests for reprints should be addressed to Dr. Jacques Ft. Caldwell, Univer- sity of Florida, 7106 Northwest 1 lth Place, Gaines- ville, Florida 32605.

In a randomized, double-blind trial, sucralfate therapy, 1 g four times daily, was compared with placebo in 143 symptomatic patients to assess the treatment of gastrointestinal symptoms and gastric mucosal damage associated with nonsteroidal anti-inflammatory drugs (NSAIDs). All patients followed a fixed regimen of NSAIDs, were assigned to one of two groups based on the presence or ab- sence of gastric erosions at baseline endoscopy, and were then assigned randomly to receive sucralfate or placebo for four weeks. Patients were then followed for up to six months while receiving open-label sucralfate 1 g twice daily to up to 1 g four times daily. After four weeks of double-blind therapy, patients taking either nonsalicylate NSAlDs or long half-life NSAlDs and who were treated with sucralfate experienced a significant reduction in both peptic symptom frequency and intensity (p CO.03) as compared with patients receiving placebo. Sucralfate-treated patients with baseline endoscopic lesions showed a significant reduction in lesion scores (p cO.005) at four weeks as compared with baseline, whereas no improvement was observed in gastric mucosal lesions of patients given placebo. Long-term sucralfate therapy resulted in continued improvement in gastrointestinal symptoms and gastric lesion scores in patients receiving all types of NSAIDs. The results indicate that sucralfate used in conjunction with NSAlDs may allow patients to continue therapy by relieving gastrointestinal symptoms and mucosal damage associated with NSAID therapy.

Upper gastrointestinal side effects caused by aspirin and other nonsteroidal anti-inflammatory drugs (NSAIDs) have compromised the usefulness of these drugs for many patients. A lack of understanding of the cause of epigastric distress and ignorance of the natural history of NSAID-induced pathology have caused most physicians and patients to discontinue therapy with NSAlDs when gastrointestinal symptoms appear. Frequently, the resultant loss of the anti-inflammatory benefits of the NSAlDs causes the underlying rheumatologic disease to flare and increases the patient’s disability. The frequency of this problem is significant. NSAlDs were used by more than 40 million persons in the United States, and nonsalicylate NSAlDs accounted for more than 4 percent, or 66.5 million, of the total prescriptions dispensed by retail pharmacies in 1983 [l].

As a result, a search for an effective method of controlling NSAID- induced gastrointestinal symptoms and irritation while maintaining the therapeutic benefits of these drugs was initiated. Many physicians pre- scribe antacids for relief of symptoms, but it remains unclear whether NSAID-induced gastroduodenal disease is acid-dependent or whether

74 September 28, 1987 The American Journal of Medicine Volume 83 (suppl 38)

SYMPOSIUM ON SUCRALFATE-CALDWELL ET AL

antacids are effective in its control. Studies of the hista- mine (Hz)-receptor antagonist cimetidine in NSAID- induced gastritis have been equivocal [2-51. Therefore, sucralfate, a gastric mucosal cytoprotective agent that has been shown to protect the gastroduodenal mucosa against the harmful effects of aspirin [6] and alcohol [7] in humans and against a variety of experimental irritants in animals [8-l I], was selected for study.

We thus undertook this double-blind, multicenter study to evaluate sucralfate’s efficacy and safety when compared with placebo in the short-term treatment of gastric mucosal damage and gastrointestinal symptoms associated with the administration of NSAIDs. A long-term open- label follow-up was added to assess any beneficial effect of sucralfate over time.

PATIENTS AND METHODS

Patients with a variety of rheumatologic disorders and who experienced upper gastrointestinal side effects while receiving a fixed daily regimen of NSAID therapy qualified for entry into this study. Patients with other significant diseases at time of entry, with endoscopically verified lower esophageal pathology, with major surgery, trauma, or burns in the preceding four weeks or during the trial, biopsy-proven carcinoma, prior gastroduodenal surgery (except for simple closure of a perforation), or addiction to ethanol or other drugs were excluded. Treatment with corticosteroids in doses exceeding 7.5 mg per day of prednisone or its equivalent, narcotics (except for pro- poxyphene), anticholinergic agents, Hz-antagonists, antacids, or investigational drugs was prohibited. Study Design. The study consisted of a two-week, single- blind, placebo-only phase followed by a four-week, double- blind placebo-controlled phase with an optional open-label sucralfate-only follow-up of up to six months. After an initial screening, patients entered a single-blind phase lasting up to two weeks, during which they received a placebo tablet four times a day while maintaining a fixed daily NSAID regimen. Six symptoms of gastrointestinal distress were followed (mentioned at the end of this paragraph); both the frequency of symptom episodes and the intensity of the most severe symptom episode experienced each day were recorded in a diary. A gastrointestinal symptom baseline was established, and the symptoms were arbitrarily divided into two catego- ries: (1) peptic: heartburn, epigastric pain/distress, epigastric burning; and (2) nonpeptic: feelings of fullness (belching, bloating, early fullness and distention, loss of appetite), nauseavomiting, cramping/diarrhea.

Patients who became asymptomatic during the single-blind phase were withdrawn from further study. Only symptomatic patients continued to participate. These patients were evaluated by upper gastrointestinal endoscopy before entering a four-week, double-blind, placebo-controlled phase. The endoscopic rating scale of the status of the gastric mucosa is shown in Table I. Symptomatic patients exhibiting gastric erosion(s) or ulceration(s) (at least grade 2) were stratified to group A, while those patients with gastrointestinal symptoms but no gastric mucosal damage (grade 0 or 1) comprised group B. For the purposes of this study, any lesion less than

TABLE I Endoscopic Rating Scale of Gastric Mucosal Injury

Group Grade Abnormalities

Group B 0 Mucosa completely normal; no injuries 1 Edema, erythema, friability, zero or one

submucosal hemorrhage Group A 2 Grade 1 and no more than three erosions

(shallow l- to 5-mm break in the mucosal surface); possibly more than one submucosal hemorrhage

3 Grade 2 or more than three erosions with evidence of several submucosal hemorrhages or inflamed gastric mucosa

4 Evidence of definite single ulceration 5 Multiple ulcers

an erosion (erythema, friability) was judged as normal gastric mucosa. Patients within each group were then randomly assigned to receive either sucralfate as a 1 -g tablet, or an iden- tical placebo tablet four times a day (one hour before meals and at bedtime) in addition to their established NSAID regimen. Antacid tablets (Amphojel300 mg) were available to be taken as needed for pain. The patients’ gastrointestinal symptoms, arthritis status, side effects, and antacid use were evaluated at two and four weeks during double-blind treatment. Those patients with baseline gastric damage (group A) underwent a second endoscopy at the end of the double- blind phase. After completing double-blind therapy, patients were eligible for entry into a long-term open-label follow-up in which they received a sucralfate l-g tablet twice daily as well as antacid tablets as needed for pain. Patients were moni- tored for up to six months in the follow-up, and the sucralfate dose could be adjusted up to 1 g four times daily if necessary for symptom control. Therapy with an NSAID was continuous during the follow-up. However, the specific NSAID prescribed or its dosage regimen could be changed by the investigator. Gastrointestinal symptoms were recorded each week in a diary, and those patients with baseline damage (group A) underwent a final endoscopy at the end of the follow-up phase. Patients were seen at the end of each month during the follow-up for evaluation of gastrointestinal symptoms, adverse effects, and antacid use. Clinical laboratory assess- ment and a physical examination were performed upon entry to the study and at the conclusion of, or upon withdrawal from, the study. Joint counts, grip strengths, 50-foot walking times, duration of morning stiffness, and patients and physi- cian global evaluations were conducted during double-blind therapy to determine the status of the patient’s arthritic symptoms. Statistical Methods. Comparison of treatment differences and determination of the statistical significance (alpha = 0.05) for endoscopic results were performed by comparing mean changes in lesion scores with the baseline mean lesion score (Wilcoxon signed rank) in each treatment group and by comparing mean changes in lesion scores between treatment groups (Wilcoxon rank sum). In analyzing treatment differences and determining statistical significance (alpha = 0.05) for symptom frequency and intensity results, daily

September 28, 1987 The American Journal of Medicine Volume 83 (suppl 38) 75


TABLE II Demographics of Patients in the Double- Blind Efficacy Study

Number in group (men/women) Mean (years) age Ulcer history

Duodenal Gastric

Consumption history Alcohol Tobacco Caffeine

Sucralfate Placebo

72 (12/60) 71 (16155) 57 ? 13.5 56 t 12.7

11 15 15 13

19 31 12 19 60 55

0 2 wk 6 wk

2.60

0 Sucralfate (N=30)

0 Placebo (N=30)

-T , 2.47 -

2ti 6wk Grade A wk 2-6

‘P < 0.005 tP < 0.16 $P <0.17

Figure 1. Mean endoscopic scores for all NSAlDs at Weeks 2 and 6 for patients receiving sucralfate or placebo in group A (symptomatic patients with gastric erosions or ulcerations of at /east grade 2).

means were calculated for each patient for each of the six gastrointestinal symptoms. These means were then averaged over two symptom groups: peptic and nonpeptic. Fi- nally, individual daily means for each symptom group were averaged to calculate a treatment group daily mean score for peptic and nonpeptic symptoms. Group daily mean scores were then compared between treatments (Wilcoxon rank sum).

RESULTS

Of the 229 patients entered into the study, 67 patients were excluded after the placebo lead-in phase because their gastrointestinal symptoms disappeared. Nineteen of the remaining 162 patients were excluded from the efficacy analysis because of protocol violations-seven in the sucralfate group and 12 in the placebo group. Of the 143 patients evaluated after double-blind treatment, 72 were assigned to sucralfate therapy and 71 to treatment with placebo. Sixty of these 143 patients had significant gastric mucosal lesions (group A). Of those 60, 30 were assigned to sucralfate therapy and 30 to treatment with placebo. The remaining 83 patients were symptomatic but

76 September 26, 1967 The American Journal of Medicine Volume 63 (suppl 3B)

had no gastric lesions (group B). Of those 83, 42 were assigned to sucralfate therapy and 41 to treatment with placebo. One hundred seven of the original 229 patients entered the long-term follow-up. Thirteen were excluded from the efficacy analysis because of protocol violations, leaving 94 patients available for the safety and efficacy analysis in the long-term follow-up. Forty-six of these 94 patients had received double-blind sucralfate while 48 had received double-blind placebo. There were no significant differences in ratios of age or sex, ulcer pattern, or history of alcohol, tobacco, or caffeine consumption between the treatment groups (Table II). NSAlDs taken and the number of patients receiving each were as follows: naproxen, 30; piroxicam, 25; diflunisal, 24; aspirin, 19; salsalate, 18; ibuprofen, 12; sulindac, 9; and other, 18. The level of antacid consumption was similar in both groups. Endoscopy. Gastric mucosal lesions were shown in 60 patients, or 42 percent, at baseline; 83 patients had symptoms but no lesions. The mean baseline endoscopic score of 3.50 f 0.20 (mean -C SEM) for the 30 group A patients receiving sucralfate was significantly higher (p ~0.05) than the score of 2.90 + 0.22 for the 30 group A patients receiving placebo (Figure 1). This statistical imbalance may represent a bias in favor of the placebo group, in which the patients began double-blind treatment with less gastric mucosal damage than those in the sucralfate group. Sucralfate-treated patients showed a significant (p ~0.005) mean decrease of 1.04 + 0.29 in the endoscopic score compared with baseline after four weeks of therapy. The placebo group showed a mean decrease of 0.43 -+ 0.28, but the difference was not significant. How- ever, a direct comparison of the mean decrease in score for patients treated with sucralfate versus placebo was not significant.

Table III shows the results of the stratified analysis of lesion score changes in patients receiving salicylates (aspirin, choline magnesium trisalicylate, and salsalate) and nonsalicylates (all other NSAIDs). A significant reduction (p CO.01) in mean lesion score of 0.96 ? 0.32 was shown in patients in the nonsalicylate group receiving sucralfate. The mean reduction of 0.14 r 0.32 in the placebo group was not statistically significant. A direct comparison of these mean decreases neared statistical significance (p ~0.08). Analysis of results in the salicylate group is limited because of the low number of patients. However, no clear advantage was found with either sucralfate or placebo treatment in patients receiving salicylates.

An analysis of patients receiving NSAlDs with a long half-life (diflunisal, naproxen, piroxicam, and sulindac) or a short half-life (aspirin, choline magnesium trisalicylate, fenoprofen, ibuprofen, indomethacin, meclofenamate, tol- metin, and salsalate) is shown in Table Ill. Sucralfate significantly (p ~0.04) reduced the mean lesion score by 0.84 -C 0.34 from baseline in the long half-life group. The mean change with placebo (0.18 ? 0.32) was not signifi-


cant. The mean changes in the sucralfate and placebo groups were not found to be significantly different (p =0.22) from one another when compared directly.

In the short half-life group, the mean changes in lesion scores in the sucralfate and placebo group were not significant. However, a reduction of 1.30 2 0.60 in the sucralfate group approached significance (p ~0.08). As in the analysis of salicylates, this analysis is limited because of the low number of patients.

Thirty-six of the original 60 patients receiving all types of NSAlDs and with baseline endoscopic lesions were moni- tored in the long-term follow-up; their data are presented in Figure 2. The mean two-week baseline score of 3.33 r 1.07 improved to 2.31 t 1.53 after four weeks of double-blind therapy (Week 6) and to 2.03 ? 1.46 at the end of long-term therapy. The decrease of 1.30 + 1.53 in the two weeks to the end of long-term therapy was statistically significant (p =O.OOOl) when compared with the baseline score. The 0.28 2 1.94 decrease from Week 6 to the end of long-term therapy was not significant (p =0.3880). As Figure 2 shows, this improvement was seen in patients formerly receiving both double-blind sucralfate and those formerly given placebo, although the change in lesion score from Week 6 to the end of the long-term period worsened (0.26 ? 1.73) but was not statistically significant (p =0.47111) for patients formerly given double-blind sucralfate. Gastrointestinal Symptoms. Peptic: Statistically significant (p ~0.05) reductions in the mean frequency and intensity of peptic symptoms (heartburn, epigastric pain/ distress, epigastric burning) were consistently found in favor of sucralfate versus placebo in the double-blind treatment phase. The advantage of sucralfate treatment was shown in reductions in both mean frequency and intensity of symptoms in patients receiving long half-life and nonsalicylate NSAlDs as well as in mean intensity reduction in those patients receiving all NSAIDs. Figures 3 to 5 show the results for peptic symptoms in patients receiving all NSAIDs, nonsalicylate NSAIDs, and long half-life NSAIDs, respectively. Patients receiving salicylates and short half-life NSAlDs showed no significant differences in favor of sucralfate or placebo therapy in reducing mean symptom frequency or intensity. Sucralfate significantly reduced the frequency and intensity of symptoms in patients receiving nonsalicylate and/or long-acting NSAIDs. However, if one examines the data for all NSAIDs, a mean decrease in frequency score of 0.36 t 0.07 was shown in the 72 patients receiving sucralfate, compared with a 0.19 + 0.07 mean decrease in frequency score in the placebo group (n = 71). A comparison of these decreases did not reach statistical significance (p ~0.12). However, the decrease in symptom intensity of 0.81 -c 0.14 at post- treatment in the sucralfate group was significantly different from the mean decrease of 0.55 + 0.13 in the placebo group (p CO.025). Thus, sucralfate therapy was more ef-



4.0

1 3.33 3.37

l! 1 I 7 *

N=36 N.M N=36 N-49 N.19 N-IS N=17 N=17 N=li

2wk 6wk LT 2 wk 6 wk LT 2 wk bwk LT

All Futients Former DB-SCR Potlents Former DBPL Patients

‘P = o.ooo1 IP = 0.008 tP = 0.006

“NS *lNS ttNS

2 wk = Basellne 6wk 5 Post Double-Blind LT = Post Long-Term

Mean Daily

score 2.00

1 Frequency

0.0 2wk 6wk 2wk Bwk GradeAwk2-3

P < 0.03

.l :

Intensity

m Sucralfafe (N-56) 1.34

1 0 Placebo (N-50)

!wk 3wk GradeAwk28 P < a03

Figure 2. Mean endoscopic gastric lesion scores at Weeks 2 and 6 and at the end of long-term therapy in continuing patients in group A (symptomatic patients with gastric erosions or ulcerations of at least grade 2).

Figure 3. Mean daily peptic symptomatic scores at Weeks 2 and 6 in patients receiving all NSAIDs.

78 September 28, 1987 The American Journal of Medicine Volume 83 (suppl 3B)


Figure 4. Mean daily peptic symptomatic scores at Weeks 2 and 6 in patients receiving nonsalicylate NSAIOs.

Figure 5. Mean daily peptic symptomatic scores at Weeks 2 and 6 in patients receiving long ha/f-life NSAIOs.

Mean Daily E&Ore

2.00 1 Frequency

2wk 6wk 2 wk 6 wk Grade A wk 2-6 P < 0.125

Intensity

1.67 n Sucralfate (N-72) 1.51 - 0 Placebo (N-n)

1.12

0.61

2 wk 6 wk 2 wk 6 wk Grade Awk 2-6 P < 0025

Maan Daily SC0m

1 Frequency Intensity

P < 0.004

B SucraNata (N943) 0 Placebo (N- 42)

6wk GtadeAwk2-6 P < 0.015

1

fective than placebo in reducing the mean intensity of the most severe daily episode of peptic symptoms in patients receiving all NSAIDs.

Even more convincing evidence of sucralfate’s efficacy in reducing the frequency and intensity of peptic symptoms was found in patients receiving nonsalicylate NSAIDs. The mean decrease in frequency (0.37 + 0.06) was significantly greater (p ~0.03) in the sucralfate group (n = 56) than in the placebo group (0.19 4 0.06; n = 50). The sucralfate group also showed a significantly greater (p ~0.03) mean reduction in intensity (0.84 2 0.15) than did the placebo group (0.62 ? 0.16).

In the long-term follow-up, continued significant reductions in peptic symptom frequency (Figure 6) and intensity (Figure 7) were demonstrated. The baseline double- blind frequency score of 0.83 2 0.92 improved to 0.23 + 0.48 at the end of long-term therapy (p = 0.0001, n = 93). The decrease of 0.35 t 0.61 from the end of double-blind therapy (Week 6) to the end of long-term therapy was also significant (p = 0.0001, n = 88). The double-blind baseline intensity scores to the end of long-term intensity scores and the end of double-blind intensity scores to long-term intensity scores also showed significant improvement (p = 0.0001) (Figure 7).

Patients receiving long half-life NSAlDs showed the Nonpeptic: No statistically significant mean reduc- strongest evidence of sucralfate’s ability to reduce both tions in either frequency or intensity of nonpeptic symp- frequency and intensity of symptoms. The mean decrease toms were found in favdr of sucralfate or placebo treat- in frequency in the sucralfate group (n = 43) of 0.39 + ment during the double-blind phase. However, statistically 0.05 was significantly different (p ~0.004) from the mean significant improvements occurred in nonpeptic frequency decrease of 0.17 2 0.05 in the placebo group (n = 42). and intensity scores of patients during long-term treat- Similarly, the mean decrease in intensity in the sucralfate ment when compared with scores at pre-double-blind and group (0.92 ? 0.15) was significantly different (p ~0.02) post-double-blind therapy. The baseline frequency score from the mean decrease in the placebo group (0.55 + of 0.94 & 1.59 and the post-double-blind score of 0.63 -C 0.16). 1.62 improved to 0.17 * 0.28 at the end of long-term ther-

September 29, 1997 The American Journal ol Medlclne Volume 93 (suppl 38) 79


1 .Ol All Patlents (N = 931

O.l-

2 wk 6 wk LT 2 wk 6 wk LT Peptic Symptoms Nonpeptic Symptoms

.P 3 O.WOl (wltcoxon signed rank tee4 2 wk I Saseline 6wk = Post DoubleBlind LT = Post Long-Term

Figure 6. Mean daily peptic and nonpeptic symptom frequency scores at Weeks 2 and 6 and at the end of long-term therapy in continuing patients receiving all types of NSAlDs on a long-term basis.

apy. The respective decreases of 0.77 + 1.41 (n = 93) and 0.45 + 1.42 (n = 88) were both statistically significant (p = 0.0001) (Figure 8). Likewise, significant decreases in the long-term intensity scores (p = 0.0001) were seen when compared with scores at pre-double-blind and post- double-blind (Figure 7).

Safety: Two patients, one in the placebo group and one in the sucralfate group, withdrew from the double- blind study as a result of an adverse reaction. Nausea and vomiting developed in the patient taking sucralfate. The patient receiving placebo experienced gastrointestinal cramping, diarrhea, rash, and swollen lips. Two additional patients receiving sucralfate withdrew from the long-term follow-up because of an adverse reaction. A rash developed on the face of one patient, and a second experienced headache. Symptoms disappeared in all four patients after cessation of therapy. Although certain fluctua- tions in the results of laboratory tests were observed, none was judged to be directly related to therapy with either sucralfate or placebo in either phase of the study.

Compliance: Similar compliance was found in both treatment groups. Patients in the sucralfate group reported taking 82 percent of the dispensed study medica- tion versus 84 percent for the placebo group during the

30 September 28,1987 The American Journal of Medicine Volume 83 (suppl 38)

All Patients (N =93)

0 2 wk 6 wk LT 2 wk 6 wk LT

Peptic Symptoms Nonpeptic Symptoms

‘P = 0.0001 IWilCoxon signed rank test) 2 wk = Baseline 6 wk d Post Double-Blind LT = Post Long-Term

Figure 7. Mean daily peptic and nonpeptic symptom intensity scores of the most severe daily episode at Weeks 2 and 6 and at the end of long-term therapy in continuing patients receiving all types of NSAlDs on a long-term basis.

double-blind phase. Sucralfate compliance was approxi- mately 90 percent during the long-term follow-up.

Inflammatory condition: The status of each patient’s arthritic condition was assessed at baseline and at Weeks 2, 4, and 6 during the study. No statistically significant differences (alpha = 0.05) were found between the sucralfate and placebo treatment groups for assessments of joint tenderness/pain or swelling, grip strength, walking time, or morning stiffness. Sucralfate had no observed effect on the clinical arthritic condition.

COMMENTS

This placebo-controlled multicenter study indicates that sucralfate is effective in relieving symptoms and reducing gastric mucosal damage in patients receiving concomitant NSAID therapy. The endoscopic results for all patients with baseline gastric mucosal damage indicated that the sucralfate-treated group experienced a significant improvement (p ~0.005) in mucosal lesion scores when compared with those at baseline. This improvement was most pronounced in those patients receiving long half-life or nonsalicylate NSAIDs. Direct comparison of gastric mucosal healing failed to show statistically significant differences between the sucralfate and placebo groups. This


lack of significance may have been influenced by the presence of more pronounced gastric mucosal damage in the sucralfate group at baseline and by the possibility of a type II statistical error caused by the small number of patients (60 patients) with mucosal damage at baseline. An improved gastric mucosa was maintained during the long- term open-label follow-up. It should be noted that patients who previously received double-blind placebo showed a larger numerical endoscopic improvement during long- term sucralfate therapy than did patients previously receiving sucralfate in a double-blind manner.

Sucralfate significantly alleviated both the frequency and intensity of peptic symptoms in patients taking NSAIDs. Symptomatic improvement, like gastric mucosal healing, was greater in patients receiving nonsalicylate or long-acting NSAIDs, such as sulindac, piroxicam, diflunisal, or naproxen. Because the long half-life NSAlDs are among the most frequently prescribed drugs for arthritic patients [l], these observations are clinically relevant to a major portion of the population that receives NSAID therapy. This improvement in symptoms was continued during long-term therapy. In fact, both peptic and nonpeptic symptom frequency and intensity scores were significantly reduced for up to six months of sucralfate long-term therapy. The continued improvement was noted in the majority of patients (60 of 94 or 63.83 percent) receiving no more than sucralfate 1 g twice daily.

NSAlDs have long been associated with gastrointestinal symptoms and the development of gastric ulcers [12,13]. Of all the NSAIDs, damage induced by aspirin is the best known and most widely studied. Gastric erosions developed in 50 percent of aspirin-treated patients with chronic rheumatic disease, and 20 percent experienced gastric ulceration in one study [14]. Patients who underwent long-term aspirin therapy also experienced duodenal lesions (13 percent with erosions, 4 percent with ulcers) 1151.

A large proportion of patients reporting gastrointestinal symptoms lack endoscopic evidence of mucosal damage [15,16]. In our study, 58 percent (83 of 143 patients) of symptomatic patients exhibited a normal gastric mucosal appearance at baseline endoscopy. The mechanism by which NSAlDs produce gastrointestinal symptoms in such patients with normal mucosa deserves further study. Gas- tric mucosal lesions were observed in 42 percent (60 of 143 patients) of our symptomatic patients. The goals of treatment of upper gastrointestinal distress in patients receiving NSAlDs must be directed at relieving symptoms and at healing any gastric mucosal damage, thus reducing the risk of serious complications such as hemorrhage, obstruction, or perforation. In this study, these goals were accomplished with sucralfate 1 g four times daily even though NSAID therapy was maintained at the drug’s full dose throughout the study. This same protection was continued during long-term therapy where 59 of 94 (62.77

percent) patients continued to receive the same NSAID at the same dose while receiving sucralfate 1 g two to four times daily.

NSAID-induced gastritis may not be acid-dependent, and treating this condition with antacids or He-receptor antagonists has been shown to have very limited effec- tiveness [2-51. In one study [2], two-months of treatment with therapeutic doses of cimetidine plus antacids was only partially effective in treating aspirin-associated gastric ulcers measuring 0.5 cm or more in diameter; larger ulcers showed greater resistance to therapy. In a four- week study involving patients with rheumatoid arthritis and osteoarthritis receiving NSAlDs excluding aspirin [3], only 43 percent of cimetidine-treated gastric ulcers had healed compared with 50 percent of gastric ulcers treated with placebo.

NSAlDs insult the mucosa through a variety of means secondary to inhibition of gastrointestinal prostaglandin synthesis [17] and disruption of the gastric mucosal bar- rier [I 81 with resulting mucosal injury [19,20]. Alteration of the quality and viscosity of gastric mucus secretion by salicylates may also play a role in damage [21,22]. An alter- native method of treating gastric lesions that are secondary to NSAID therapy is to enhance the protective physiol- ogy of the gastric mucosa with a “cytoprotective” agent. Cytoprotection is a physiologic and pharmacologic phe- nomenon that is independent of acid secretion by which the mucosa is protected from deep necrosis normally induced by necrotic and ulcerogenic challenges [23,24].

Sucralfate, a sulfated polysaccharide, has been shown to be effective in the treatment of duodenal [25,26] and gastric [27] ulcer disease and was investigated in the cur- rent study because its mechanism of action involves cytoprotection of the gastric mucosa. Its cytoprotective properties have been demonstrated in several animal studies through stimulation of endogenous prostaglandin E2 synthesis [a], local prostaglandin release [28,29], in- creased mucus output [28,30,31], and enhanced mucosal proliferative zone stimulation [28,32]. In humans, sucralfate has been shown to increase mucus concentration and provide more complete protection of the gastric mucosa against alcohol-induced microvascular damage and deep gastric mucosal lesions than did cimetidine [7]. Thus, sucralfate enhances the gastric mucosal protective factors prostaglandin E2 and mucus, which are the factors NSAlDs have been shown to inhibit.

In our multicenter study, 143 patients with NSAID- induced gastrointestinal symptomatology have been ob- sewed for four weeks while receiving full-dose NSAID therapy. None of these patients developed any significant gastrointestinal complications as a result of continuing therapy with NSAIDs. In fact, just 42 percent of patients with symptoms had gastric mucosal damage that could be demonstrated at baseline endoscopy. The patients who had mucosal damage and who were treated with sucral-



fate experienced partial but not total improvement in the both peptic and nonpeptic symptom control was seen in appearance of the gastric mucosa. The data suggest that the long-term follow-up. Although the long-term observa- this improvement may be more significant in patients tions were uncontrolled, these results support the findings treated with long half-life NSAIDs, such as naproxen, pirin the four-week controlled portion of the study, and no oxicam, sulindac, and diflunisal, than in patients treated significant gastrointestinal complications arose during with short-acting NSAIDs. Improvement may also be long-term therapy. greater in patients treated with nonsalicylate than salicy- Upper gastrointestinal symptomatology associated with late NSAIDs, although further research would be neces- NSAID therapy frequently fosters discontinuation of such sary to determine the consistency of this relationship. In therapy. Our reported data do not support such a practice. addition, the intensity and frequency of peptic symptoms This study shows that patients may be allowed to continue were relieved by sucralfate. Again, improvement was receiving NSAlDs when a cytoprotective agent such as most clearly evident in those patients receiving long half- sucralfate is used as an adjunct to partially relieve both life or nonsalicylate NSAIDs. Improvement in the gastric the gastrointestinal symptoms and gastric mucosal mucosa was maintained, and continued improvement in damage associated with NSAID therapy.

REFERENCES

1.

2.

3.

4.

5.

6.

7.

6.

9.

10.

11.

12.

13.

14.

15.

16.

17.

Baum C, Kennedy DL, Forbes MB: Utilization of nonsteroidal anti-inflammatory drugs. Arthritis Rheum 1965; 28: 666-692.

O’Laughlin JC, Silvoso GK, lvey KJ: Resistance to medical therapy of gastric ulcers in rheumatic disease patients taking aspirin. A double-blind study with cimetidine and follow-up. Dig Dis Sci 1982; 27: 978-960.

Davies J St J, Dixon A, Brales JS: The effect of cimetidine on peptic ulceration in patients receiving anti-inflammatory therapy. In: Cimetidine. The Westminster Hospital Symposium. Edinburgh: Churchill-Livingston, 1978; 275-280.

Loludice TA, Saleem T, Lang JA: Cimetidine in the treatment of gastric ulcer induced by steroidal and nonsteroidal anti- inflammatory agents. Am J Gastroenterol 1981; 75: 104-l 10.

O’Laughlin JC, Silvoso GR, lvey KJ: Effects of an Hz-blocker on gastric ulcers in patients taking aspirin: a double-blind study (abstr). Gastroenterology 1980; 78: 1230.

Tesler MA, Lim ES: Protection of gastric mucosa by sucralfate from aspirin-induced erosions. J Clin Gastroenterol 1981; 3 (suppl 2): 175-179.

Tarnawski A, Glick ME, Stachura J, et al: Comparison of sucralfate vs cimetidine in protection of the human gastric mucosa against alcohol injury (abstr). Gastroenterology 1985; 88: 1609.

Ligumsky M, Karmeli F, Rachmilewitz D: Sucralfate stimulation of gastric PGE:! synthesis-possible mechanism to explain its effective cytoprotective properties (abstr). Gastroenterology 1984; 88: 1164.

Harrington SJ, Schlegel MS, Code CF: The protective effect of sucralfate on the gastric mucosa of rats. J Clin Gastroenterol 1981; 3 (suppl 2): 129-134.

Nagashima R, Hoshino E, Hinohara Y, et al: Effect of sucralfate on ethanol-induced gastric mucosal damage in the rat. Stand J Gastroenterol 1983; 16 (suppl 83): 17-20.

Okabe S, Takeuchi K, Kunimi H, et al: Effects of an antiulcer drug, sucralfate (a basic aluminum salt of sulfated disaccha- ride), on experimental gastric lesions and gastric secretion in rats. Dig Dis Sci 1983; 28: 1034-l 042.

Simon LS, Mills JA: Nonsteroidal anti-inflammatory drugs. N Eng J Med 1980; 302: 1179-l 164.

Simon LS, Milk JA: Nonsteroidal anti-inflammatory drugs. N Eng J Med 1980; 302: 1237-1243.

lvey KJ: Drugs, gastritis, and peptic ulcer. J Clin Gastroenterol 1961; 3 (suppl 2): 29-34.

Lockard 00, lvey KJ, Butt JH, et al: The prevalence of duodenal lesions in patients with rheumatic diseases on chronic aspirin therapy. Gastrointest Endosc 1980; 26: 5-7.

Binder JH, Cocco A, Crossley RJ, et al: Cimetidine in the treatment of duodenal ulcer. A multicenter double-blind study. Gastroenterology 1978; 74: 380-388.

Vane JR: inhibition of prostaglandin synthesis as a mechanism

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

of action for aspirin-like drugs. Nature New Biol 1971; 281: 232-235.

Kivilaakso E, Silen W: Pathogenesis of experimental gastromucosal injury. N Engl J Med 1979; 301: 364-369.

Lanza FL, Royer GL, Nelson RS, et al: The effects of ibuprofen, indomethacin, aspirin, naproxen and placebo on the gastric mucosa of normal volunteers. Dig Dis Sci 1979; 24: 823-827.

Lanza F, Royer G, Nelson R: An endoscopic evaluation of the effects of nonsteroidal anti-inflammatory drugs on the gastric mucosa. Gastrointest Endosc 1975; 21: 103-105.

Minguy R, Masters YF: Effects of aspirin in gastric mucus secretion. Surg Gynecol Obstet 1965; 120: 92-98.

Kent PW, Allen A: The biosynthesis of intestinal mucus. The effect of salicylates on glycoprotein biosynthesis by sheep colonic and human gastric mucosal tissue in vitro. Biochem J 1966; 108: 645-656.

Robert A, Nezamis JE, Lancaster C, Hanchar AJ: Cytoprotec- tion by prostaglandins in rats. Prevention of gastric necrosis produced by alcohol, HCI, NaOH, hypertonic NaCI, and ther- mal injury. Gastroenterology 1979; 77: 433-443.

Robert A: Cytoprotection by prostaglandins. Gastroenterology 1979; 77: 761-767.

Martin F, Forley A, Gagron M, Poitras P: Comparison of the healing capacities of sucralfate and cimetidine in the short- term treatment of duodenal ulcer: a double-blind randomized trial. Gastroenterology 1982; 82: 401-405.

Hollander D: Efficacy of sucralfate for duodenal ulcers: multicenter double-blind trial. J Clin Gastroenterol 1981; 3 (suppl 2): 153-157.

Martin F, Farley A, Gagnon M, Poitras P, Bensemana D: Short- term treatment with sucralfate or cimetidine in gastric ulcer: preliminary results of a controlled randomized trial. Stand J Gastroenterol 1983; 16 (suppl 83): 37-41.

Tarnawski A, Hollander D, Stachura J, et al: Trophic effect of long-term sucralfate on the gastric mucosa. A key to prophy- lactic efficacy? (abstr). Gastroenterology 1985; 88: 1610.

Hollander D, Tarnawski A, Gergely H, Zipser RD: Sucralfate protection of the gastric mucosa against ethanol-induced injury: prostaglandin-medicated process? Stand J Gastroen- terol 1984; 19 (suppl 101): 97-102.

Guslandi M, Ballarin E, Cambielli M, et al: Stimulation of mucus output: a further mode of action of sucralfate (abstr). Gastro- enterology 1983; 64: 1177.

Shea-Donohue T, Steel L, Montcalm E, et al: Gastric protection by sucralfate: role of mucus and prostaglandins. Gastroenter- ology (abstr). 1985; 68: 1583.

Hollander D, Tarnawski A, Krause WJ, Gergely H: Protective effect of sucralfate against alcohol-induced gastric mucosal injury in the rat. Gastroenterology 1965; 88: 366-374.

82 September 28, 1987 The American Journal of Medicine Volume 83 (suppl 38)

Documents

Sucralfate treatment of nonsteroidal anti-inflammatory drug-induced gastrointestinal symptoms and mucosal damage