16,16-Dimethyl prostaglandin E2 protects gastric mucosal surface epithelial cells from indomethacin-induced damage in rats

The protective effect of 16,16-dimethyl prostaglandin E2 (dmPGE2) against early damage induced by indomethacin in the rat gastric mucosal surface epithelial cells was studied using a scanning electron microscope. Indomethacin (10 or 25 mg/kg, p.o.) induced a widespread exfoliation of the surface epithelial cells and an exposure of the lamina propria both in the corpus and antrum within 1 hr after the administration. Pretreatment with dmPGE2 (0.3, 3 or 30 micrograms/kg, p.o.) 30 min before indomethacin (25 mg/kg) dose-dependently inhibited these damages. The effects of dmPGE2, at least on the surface epithelial cells in the corpus, appear to be related to the prevention of damage formation itself and is unrelated to the enhancement of reconstitution of once damaged mucosa. Enhanced gastric motility by indomethacin was potently inhibited by pretreatment with 3 and 30 micrograms/kg of dmPGE2, but not with 0.3 micrograms/kg. dmPGE2 pretreatment (30 micrograms/kg) significantly decreased the absorption of indomethacin (25 mg/kg) when determined 10 min after giving indomethacin, but did not affect it when determined 30 and 60 min later. We conclude that dmPGE2 protects gastric mucosal surface epithelial cells from indomethacin injury at an early stage, partly by inhibiting gastric motility.     

Effects of 16,16-dimethyl prostaglandin E2 on surface epithelial cell damage in the rat stomach induced by vagal nerve stimulation

The effect of 16,16-dimethyl prostaglandin E2 (dmPGE2) on gastric surface epithelial cell (SEC) damage induced by vagal nerve stimulation (VS) in urethane-anesthetized rats was studied using scanning electron microscopy. VS (1.25-10 Hz, 0.2 mA, 2 msec, 10 min) resulted in a graded increase in the SEC damage, increased gastric contractions, increased gastric acid secretion, and a decrease in heart rate. Pretreatment with dmPGE2 (0.3-30 micrograms/kg, s.c.) significantly protected the SEC from VS-induced damage, inhibited the increase in gastric contractions and acid secretion, but had no influence on the decrease in heart rate. Atropine (1 mg/kg, s.c.) also protected the SEC from VS-induced damage and inhibited the alterations in response to VS. Timoprazole (30 mg/kg, s.c.) had no protective effects on SEC from VS-induced damage, no effect on increased gastric contractions and heart rate, but did inhibit the increase in gastric acid secretion, in response to VS. These results suggest that: VS-induced SEC damage was caused by increased gastric contractions and not by increased gastric acid secretion, and dmPGE2 protects against SEC damage by inhibiting gastric contractions.     

Pathogenic mechanisms involved in mepirizole-induced duodenal damage in the rat

Mepirizole (60 and 200 mg/kg) administered s.c. induced damage in the surface epithelial cells of the rat proximal duodenum as early as 2 hr after the treatment. 16,16-Dimethyl prostaglandin E2 (dmPGE2, 30 micrograms/kg) administered s.c. significantly protected the duodenal mucosa against mepirizole-induced damage for up to 6 hr. Gastric acid secretion in acute fistula preparations was significantly reduced 1 hr after administration of mepirizole (60 and 200 mg/kg). The secretion reverted to the control level 2 hr later. In the 60 mg/kg-treated group, however, there was a significant increase in the acid output for up to 6 hr. Duodenal HCO3- secretion, stimulated with 10 mM HCl was significantly inhibited with mepirizole (60 and 200 mg/kg). Mepirizole (60 and 200 mg/kg) significantly increased the amount of acid in the duodenum for 2 to 6 hr after the treatment. dmPGE2 (30 micrograms/kg) significantly inhibited gastric acid secretion, stimulated duodenal HCO3- secretion, and reduced the increased amount of acid in the duodenum in response to mepirizole. Endogenous prostaglandin E2 and 6-keto prostaglandin F1 alpha in the duodenal mucosa were significantly reduced by mepirizole (200 mg/kg) 1 to 2 hr later. Mepirizole-induced duodenal damage appears to be caused by the increased amount of acid in the duodenum.     

Effects of cold-restraint stress on gastric ulceration and motility in rats

The effects of stress by restraint at 4 degrees C for 2 hr and of drug treatment on gastric lesion formation and motor activities (contraction frequency, amplitude and tone) were studied in rats. Restraint at room temperature (22 degrees C) produced a small ulcer index in the controls and did not significantly affect gastric motor activities; atropine and verapamil reduced but bethanechol increased gastric contractions under the same experimental conditions. Restraint at 4 degrees C markedly elevated the ulcer index. The frequency of gastric contractions was significantly increased in the first hr but the amplitude was depressed during the whole 2-hr observation period. Gastric tone initially fell but rose in the second hr of cold-restraint stress. Atropine and verapamil pretreatment prevented stress-induced ulcer formation and suppressed the frequency and amplitude of gastric contraction. Bethanechol stimulated both frequency and amplitude without significantly influencing stress ulcer size. It is unlikely that gastric hypermotility plays a major role in stress ulceration; the stomach smooth muscle-relaxing action of atropine and verapamil may contribute only partly to their antiulcer effects.     

Effects of antimuscarinic agents and prostaglandin E2 on the gastric mucosal lesions induced by necrotizing agents and water-immersion stress in rats

The role of antimuscarinic action in gastric mucosal protection against necrotizing agents and the role of such mucosal protection in antiulcerogenic action were studied in rats with i.v. administered antimuscarinic agents. Pirenzepine, as well as PGE2, prevented the gastric mucosal lesions induced by all necrotizing agents (99.5% ethanol, 0.6 N HCI, 0.15 N NaOH, 0.4 N HCI-50 mM taurocholate), but atropine did not prevent the HCI-induced lesions. Cimetidine inhibited only the ethanol-induced lesions even at the antisecretory dose. Higher doses of pirenzepine (5-fold) and atropine (10-fold) were required to inhibit the gastric secretion in Shay rats than in vagally stimulated rats. There was no difference between the antisecretory doses of cimetidine in Shay rats and vagally stimulated rats. PGE2 (0.03-0.1 mg/kg) did not affect gastric secretion. The protective doses of pirenzepine and atropine against mucosal lesions induced by necrotizing agents were similar to the dose in inhibiting vagally stimulated acid secretion and water-immersion stress-induced lesions. PGE2 (100 micrograms/kg) did not prevent the water-immersion stress induced gastric lesions. These results suggested that antimuscarinic agents protect the gastric mucosa from necrotizing agents via a blocking action on the activation of the intrinsic cholinergic nerve. However, antiulcerogenic action is more deeply concerned with antisecretory action than cytoprotection.     

Mechanisms involved in aggravation of ethanol-induced gastric mucosal lesions in adrenalectomized rats

Effects of adrenalectomy (AD) on ethanol-induced gastric injury and prostaglandin (PG) protection on the damage were investigated in rats and compared with those of N-ethylmaleimide (NEM), a sulfhydryl (SH) blocker, and diethyl maleate (DEM), a SH depletor. Oral administration of 100% ethanol (1 ml) induced elongated bands of hemorrhagic lesions in the corpus mucosa of sham operated rats, and these lesions were significantly prevented by 16,16-dimethyl PGE2 (dmPGE2, 10 micrograms/kg, s.c.). AD markedly enhanced the mucosal ulcerogenic responses caused by ethanol and abolished the protective effect of dmPGE2; this agent rather worsened the lesions, which appeared throughout the corpus mucosa. AD by itself enhanced the microvascular permeability in the gastric mucosa without any effect on SH contents. These alterations caused by AD were significantly reverted by hydrocortisone treatment (10 mg/kg/day for 2 weeks, s.c.). On the other hand, a single injection of NEM (10 mg/kg, s.c.) similarly enhanced the vascular permeability, worsened the ethanol-induced lesion, and mitigated the protective effect of dmPGE2 without altering mucosal SH contents, while DEM (1 ml/kg, s.c.) significantly reduced the mucosal SH levels and the lesions. These results suggest that AD worsened the mucosal lesions induced by ethanol, probably by enhancing the microvascular permeability, and this action may be due to a lack of steroid secretion but is not directly related to a mucosal SH deficiency.     

Effects of troxipide on acute gastric lesions in rats

Effects of troxipide on several acute gastric lesions in rats were investigated in comparison with those of cetraxate. Troxipide (100, 200, 300 mg/kg) and cetraxate (100, 300, 1,000 mg/kg), given orally, dose-dependently protected the gastric mucosa from damage due to ethanol. Aspirin- and 0.6 N HCl-induced gastric lesions were dose-dependently inhibited by troxipide (200, 300 mg/kg), but only significantly inhibited by cetraxate at high dose (1,000 mg/kg). Troxipide (100, 200, 300 mg/kg) dose-dependently prevented the formation of gastric lesions induced by water-immersion stress, whereas cetraxate (600, 1,000 mg/kg) also significantly prevented gastric lesions. That is, protective effects of troxipide were much more potent than those of cetraxate against aspirin-, 0.6 N HCl- and water-immersion stress-induced gastric lesions, whereas both were almost equal against ethanol-induced gastric lesions. In addition, cytoprotective effects of troxipide against ethanol-induced lesions were most remarkable at 10, 30, 60 min after administration (100, 300 mg/kg) and lasted for up to 240 min. These results suggested that troxipide might be useful for the treatment of acute gastric lesions in humans.     

Gastric mucosal protection induced by restraint and water-immersion stress in rats.

We found that the gastric mucosal lesion induced by 60% ethanol containing 150 mM HCl was reduced by pre-loading the rat with restraint and water-immersion stress (22 degrees C). The resistance to ethanol injury was maximal in 1-hr stress loaded rats and gradually decreased with increasing time of stress (2-6 hr). The protective effect of stress was markedly decreased by subdiaphragmatic truncal vagotomy, and electrical stimulation of vagal nerves afforded similar protection as observed after stress. In contrast, pretreatment with atropine markedly reduced ethanol injury in both the control and 1 hr stress-loaded rats. One-hour stress produced surface epithelial cell damage in sham vagotomized rats, but the surface cells were almost intact in vagotomized rats. Pretreatment with indomethacin significantly mitigated the protective effect of stress against ethanol injury. However, the level of mucosal prostaglandin E2 was not changed 1 hr after stress, and it tended to decrease 6 hr after stress. Pretreatments with 1-hr stress and vagal stimulation weakly reduced localized staining of gastric mucosa with gentian violet. We conclude that adaptive protection can be achieved by restraint and water-immersion stress.     

Dissolution of antisecretory and cytoprotective action of PGE2 in rats

Oral administration of 60% ethanol in 150 mM HCl (HCl.ethanol) to rats produced gastric mucosal lesions within 1 hr. PGE2 given s.c. at 0.1 to 3 mg/kg 0.5 hr before HCl.ethanol significantly prevented the lesion formation. The protection afforded with 3 mg/kg of PGE2 persisted for 6 or 12 hr after administration. PGE2 given s.c. at 3 mg/kg significantly inhibited gastric secretion in pylorus-ligated and intact rats, but the action disappeared 6 hr later.     

Effect of 16,16-dimethyl prostaglandin E2 on gastric surface epithelial cell damage induced by 20% ethanol in rats

Prostaglandins protect against the gross damage of gastric mucosa induced by 50-100% ethanol, but do not protect surface epithelial cells (SEC) from necrosis. Since this induced damage to SEC is so severe, we attempted to determine the effects of a prostaglandin on slightly induced SEC damage and gastric potential difference (PD) in response to low concentrations of ethanol. The necrotizing effects of graded concentrations of ethanol (10-50%) to SEC on the rat gastric mucosa were studied by scanning electron and light microscopy. Intragastric instillation of 20% ethanol (v/v, 1 ml/100 g body wt.) to pylorus-ligated rats for 10 min induced slight and reproducible SEC damage consisting mainly of the apical cell membrane erosion of SEC. Pretreatment with 16,16-dimethyl prostaglandin E2 (dmPGE2, 3 or 30 micrograms/kg, p.o. or s.c.) afforded protection of the SEC from 20% ethanol-induced damage. However, the cytoprotective effects of dmPGE2 were abolished when gastric contents were emptied prior to 20% ethanol instillation. Intragastric instillation of ethanol immediately reduced PD in a concentration-related manner. dmPGE2 (3 or 30 micrograms/kg, s.c.) had no effect on the reduction of gastric PD after 20% ethanol treatment and the recovery of reduced PD to normal levels. We conclude that dmPGE2 has no cytoprotective effect on 20% ethanol-induced SEC damage in rat gastric mucosa.     

Role of gastric acid and bile acids in the pathogenesis of compound 48/80-induced gastric lesions in rats

We studied the effects of various agents, which influence gastric acidity and bile acids, on compound 48/80 (48/80)-induced gastric lesions in rats. 48/80-Induced gastric lesions were produced by repeated intraperitoneal administration of 48/80 at 0.75 mg/kg once daily for 4 days. Test agents were given orally twice daily (30 min before and 9 hr after 48/80 administration) for 4 days. AI(OH)3 and sucralfate at 2000 mg/kg/day, a weak antacid dose, significantly inhibited (about 50-60%) the development of 48/80-induced lesions. Propantheline at 60 mg/kg/day and omeprazole at 60 or 200 mg/kg/day, which reduced gastric secretion for more than 12 hr, also significantly inhibited (about 30-40%) these lesions. Cimetidine at 200 mg/kg/day, which reduced gastric secretion for only 5 hr, had little effect on the lesion formation. Cholestyramine, which is a potent bile acids binding agent, had no effect on 48/80-induced lesions in doses of 600 or 2000 mg/kg/day. These results suggest that gastric acid, but not bile acids, is partly involved in the pathogenesis of 48/80-induced gastric lesions.     

Sulphasalazine and experimental stress ulcers

The effects of sulphasalazine on gastric ulceration induced by restraint at 4 degrees C (stress) for 2 h were studied in rats. Doses of 63 or 125 mg/kg s.c., which had no effect on stomach wall prostaglandin E2 (PGE2) levels, prevented stress ulceration but not the lesions produced by indomethacin. Stress significantly increased gastric glandular mucosal PGE2 levels. Indomethacin pretreatment (20 mg/kg, p.o.) markedly reduced PGE2 levels in the same region of the stomachs, and worsened stress-induced lesion formation. Pretreatment with sulphasalazine of animals given indomethacin and then subjected to stress did not appear to affect the indomethacin component of indomethacin-stress ulceration. Oral administration of PGE2 200 micrograms/kg significantly elevated gastric PGE2 levels, but had no effect on stress ulceration. It appears that neither the antiulcer activity of sulphasalazine nor stress-induced ulceration is associated with gastric tissue PGE2 increase or decrease, respectively. The protective mechanism may result from the ability of sulphasalazine to inhibit lipoxygenase activity.     

Effects of misoprostol, (+/-)-methyl (11 alpha, 13E)-11, 16-dihydroxy-16-methyl-9-oxoprost-13-en-l-oate, on various gastric and duodenal lesions in rats

Male Sprague-Dawley rats (230-280 g), either fasted for 15-24 hr or non-fasted prior to experiments, were used. Misoprostol (3-100 micrograms/kg, p.o.) dose-dependently inhibited the development of 150 mM HCl X aspirin (100 mg/kg)-, 150 mM HCl X 60% ethanol-, and aspirin (150 mg/kg)-induced gastric lesions. Misoprostol (30, 100 micrograms/kg, p.o.), given twice daily for 4 days, significantly inhibited prednisolone (50 mg/kg given once daily for 4 days)-induced gastric lesions. Misoprostol (30 or 2 X 300 micrograms/kg, p.o.) also significantly inhibited water-immersion stress (21 degrees C, 10 hr)-induced gastric lesions or mepirizole (200 mg/kg)-induced duodenal lesions, respectively. In contrast, misoprostol (30-300 micrograms/kg, p.o.) had no effects on indomethacin (25 mg/kg)- and mepirizole (200 mg/kg)-induced gastric lesions. Misoprostol (30 micrograms/kg, p.o.) had no effect on gastric secretion in pylorus-ligated preparations (4 hr), but it (100 or 300 micrograms/kg, p.o.) significantly increased the volume and pepsin output. Gastric motility, either normal or enhanced with indomethacin (25 mg/kg), was inhibited by misoprostol (30 or 300 micrograms/kg, p.o.). Misoprostol (30 micrograms/kg, i.d.) significantly stimulated duodenal HCO3- secretion. Mechanisms by which misoprostol inhibits various gastric lesions remain unknown. However, the stimulatory activity on duodenal HCO3- secretion appears to be involved in the preventive effect of misoprostol on the development of duodenal lesions. The effects of cimetidine and 16,16-dimethyl PGE2 were also studied and compared with those of misoprostol.     

Effects of NIK-228 on gastric acid secretion in rats using the congo red sprayed method]

We have reported the antiulcer activities of a new compound that we named NIK-228 (3-hydroxy-methyl-2-methylimidazo [2, 1-b] benzothiazole). In the present report, we studied the antisecretory effects of NIK-228 on basal and stimulated gastric acid secretion using the Congo red sprayed method. Male Wistar rats (200 to 250 g) were used after 24 hr of fasting (without water). NIK-228, atropine and cimetidine were administered orally or intravenously 1 hr before operation for Congo red spraying. NIK-228 (100 mg/kg, p.o.), atropine (5 mg/kg p.o.) and cimetidine (100 mg/kg, p.o.) all inhibited basal gastric acid secretion. Oral administration of NIK-228 and atropine inhibited gastrin, 2-deoxy-D-glucose (2-DG) and bethanechol-induced acid secretion, but didn't inhibit histamine-induced acid secretion. Cimetidine inhibited all of histamine, gastrin, 2-DG and bethanechol-induced acid secretion. In vagotomized rats, oral and intravenous administration of atropine both inhibited bethanechol-induced acid secretion, but NIK-228 was not inhibited. These results suggested that antisecretory effects of NIK-228 were caused by the central vagal systems.     

Preventive actions of N-(3-aminopropionyl)-L-histidinato zinc (Z-103) through increases in the activities of oxygen-derived free radical scavenging enzymes in the gastric mucosa on ethanol-induced gastric mucosal damage in rats.

Z-103 at 1 to 25 mg/kg, p.o. prevented 100% ethanol-induced gastric mucosal lesions in a dose-dependent manner. Z-103 at 3 to 25 mg/kg, p.o. significantly elevated gastric mucosal superoxide dismutase (SOD)-like activity 1 hr after its administration to normal rats. In addition, Z-103 at doses (10 and 25 mg/kg, p.o.) which prevented 100% ethanol-induced gastric lesion further increased gastric mucosal SOD-like and glutathione peroxidase (GSH-px) activities elevated by 60% ethanol. Z-103 (10 and 25 mg/kg) significantly inhibited the increase in thiobarbituric acid-reactive substances in gastric mucosa injured by 60% ethanol. The combination with cycloheximide, a protein synthesis inhibitor, completely abolished the prevention of 60% ethanol-induced gastric mucosal lesions and the elevation of both free radical scavenging enzyme activities in the mucosa by Z-103 (10 mg/kg, p.o.). These results suggest that Z-103 may partly protect rat gastric mucosa against ethanol-induced damage by scavenging oxygen-derived free radicals via increases in the synthesis of SOD-like and GSH-px enzymes in the mucosa.     

The effect of cold-restraint stress on gastric emptying in rats

The effects of drug treatment and of cold-restraint stress (a method used to produce experimental stomach ulcers) on gastric emptying of a resin (colestipol-phenol red complex) were investigated in rats. Gastric emptying was decreased by intraperitoneal treatment with atropine (0.3 mg/kg) or verapamil (4 mg/kg), and enhanced by bethanechol (1.2 mg/kg). Stress by restraint at 4 degrees C for 2 hr markedly reduced gastric emptying; the pattern of effects of drug pretreatment in these stressed rats was similar to that seen in their nonstressed controls. Further experiments, with stress for 3 hr, revealed that the gastric emptying rate was triphasic; increasing in the first hr, returning to normal and then slowing in the third hr of stress. Initial increase in emptying rate was probably due to predominant vagal overactivity. Hypothermia and possibly other factors induced by cold-restraint stress could have subsequently depressed gastric motility.     

Histamine-induced villous damage in the rat duodenum

A single s.c. administration of histamine dose-dependently (5-20 mg/kg) induced villous damage of the proximal duodenum in 24-hr fasting rats. Time course studies indicate that histamine (20 mg/kg) induced severe exfoliation of the epithelial cells at the villous tips of the duodenal mucosa 0.5 hr after administration. The damage, however, tended to heal with time, and recovery was nearly complete 8 hr later. This villous damage was significantly inhibited by pretreatment with sodium bicarbonate given orally or cimetidine, omeprazole and NC-1300 given subcutaneously. Histamine (20 mg/kg) significantly stimulated gastric acid secretion and lowered the intraduodenal pH for 1 hr. Gastric content was significantly greater than that in the control group for 1 hr after histamine administration, probably due to stimulated gastric secretion and delayed emptying. We conclude that a single administration of histamine induces microscopical duodenal damage by stimulation of gastric acid secretion, but the damage heals with time, probably as a result of the short periods of acid stimulation and delayed emptying.     

Effects of 3-hydroxymethyl-2-methylimidazo [2, 1-b] benzothiazole (NIK-228) on gastric acid secretion and various experimental peptic ulcers in rats

We examined the antisecretory and antiulcer activities of NIK-228 in rats. Male Wistar rats (200 to 250 g) were used under 24 to 48 hr fasted (without water) conditions. NIK-228 and famotidine were administered orally 1 hr before pylorus ligation, stress or each ulceration inducer. Both NIK-228 (10 to 100 mg/kg) and famotidine (0.3 to 3 mg/kg) dose-dependently inhibited gastric secretion in pylorus ligated rats. Water-immersion stress-, indomethacin- or pylorus ligation (Shay)-induced gastric ulcers were dose-dependently inhibited by NIK-228 (10 to 100 mg/kg), but only water-immersion stress and indomethacin induced ulcers were dose-dependently inhibited by famotidine (0.03 to 3 mg/kg). Ethanol- and 0.6 N HCl-induced gastric lesions were remarkably inhibited by NIK-228 (ED50 = 2.7 and 5.6 mg/kg), but tended to be inhibited also by famotidine (0.3 to 3 mg/kg). Cysteamine-induced duodenal ulcer was inhibited significantly by NIK-228 (30, 100 mg/kg) or famotidine (3 mg/kg). NIK-228 may produce its antiulcer effects via antisecretory and cytoprotective effects. These results suggest that NIK-228 has antisecretory and antiulcer activities.     

Inhibitors of gastric lesions in the rat.

The production by stress of gastric lesions in rats was inhibited by metiamide and by mepyramine. Lesions induced by indomethacin treatment were inhibited by mepyramine but not by metiamide. Those induced by aspirin treatment in pylorus-ligated rats were not affected by either antihistamine drug. Oral glutamine inhibited lesion production in all three systems whereas aspirin orally markedly potentiated it. Sodium salicylate inhibited both indomethacin-induced lesions and those produced by aspirin in pylorus-ligated rats. On the other hand, copper salicylate inhibited stress-induced lesions and it, like copper aspirinate, also markedly reduced the extent of lesions produced by aspirin. On the basis of these results, stress-induced lesion production offers a suitable animal model for testing anti-ulcer drugs as it is, like most human gastric ulcers, inhibited by H2-receptor inhibitors like metiamide.     

The effects of sulindac and its metabolites on acute stress-induced gastric ulcers in rats

Rats were given a single intragastric administration of the prodrug sulindac (4.0 mg/kg) or its sulfide (1.0, 2.0, 4.0, or 8.0 mg/kg) or sulfone (1.0, 2.0, 4.0, or 8.0 mg/kg) metabolites and were then subjected to acute stress in the form of immobilization for 3 hr in a cold environment. Control rats received an equal volume of propylene glycol vehicle or nothing po. Other rats received 200 mg/kg acetylsalicylic acid (ASA) with or without stress, to compare the gastrointestinal effects of sulindac metabolites with those of a known non-steroidal anti-inflammatory agent. The sulfide metabolite exacerbated stress-induced gastric glandular ulcer incidence and severity in a dose-related manner relative to all groups except the ASA-stress group, which exhibited the greatest amount of gastric damage. The sulfone metabolite did not potentiate ulcer incidence or severity beyond control (stress only) levels at lower doses. However, at 4.0 and 8.0 mg/kg, the observed ulceration was greater than that seen in stressed but otherwise untreated animals. Sulindac, vehicle, and otherwise untreated rats exhibited a similar degree of stress-induced gastric damage. It appears that the prodrug does not significantly enhance stress-related gut disease, but that the active sulfide metabolite does. Although the clinical literature suggests that the sulfone metabolite is inactive, the present results suggest otherwise. While this metabolite did not, by itself, induce gastric damage at higher doses, sulfone did exacerbate stress ulcer formation. This is the only report of which we are aware, indicating a possible toxic effect of the sulfone metabolite.