Nocturnal Metiamide Treatment in the Management of Healed Duodenal Ulceration

Gut: first published as 10.1136/gut.18.6.438 on 1 June 1977. Downloaded from

Gut, 1977, 18, 438-441

Nocturnal metiamide treatment in the management of healed duodenal ulceration

M. H. THOMPSON, C. W. VENABLES, P. A. SMITH, AND W. WALKER From the Department ofSurgery, Clinical Biochemistry, and Haematology, Royal Victoria Infirmary, Newcastle upon Tyne

SUMMARY This paper presents the results of a pilot study to investigate whether the administration of a nocturnal dose of metiamide (the first orally active H2 receptor antagonist) would prevent or delay the relapse of duodenal ulceration after initial ulcer healing. Sixteen patients took part in a double-blind trial to compare metiamide (400 mg) with placebo. Endoscopically confirmed duodenal ulcer relapses occurred in two out of eight on metiamide and six out of eight on placebo. There was a significant prolongation of remission in those on the active drug with an apparent reduction in duodenitis.

The histamine H2-receptor antagonist metiamide Methods inhibits gastric secretion and has been shown to be an effective therapeutic agent in duodenal ulcer Seventeen ambulant hospital outpatients were disease: its administration over a 28-day period entered: 13 had duodenal ulcers and four recurrent results in ulcer healing in the majority of cases ulcers after previous surgery for duodenal ulcers. http://gut.bmj.com/ (Lancet, 1975; Thompson et al., 1975). If treatment All patients were treated initially in a double-blind is stopped after this period a significant number of trial of metiamide according to the protocol pre- patients may relapse in the ensuing weeks: only two viously described (Thompson et al., 1975). Those out ofthe 14 patients reported previously (Thompson patients whose ulcers failed to heal after admini- et al., 1975) remained symptom-free in the ensuing stration of placebo were offered 28-days' treatment year. We, therefore, felt that some kind of mainten- with metiamide, according to the same basic protocol ance treatment was desirable. Both haematological as the initial trial, endoscopy being repeated on day on September 26, 2021 by guest. Protected copyright. (Forrest et al., 1975) and mild biochemical (Thomp- 28. Thus all patients entering the present trial had son et al., 1975) side-effects of metiamide have been endoscopically confirmed healing of their duodenal reported and so continued treatment with doses of ulcer. These 17 patients were then re-randomised for the order of 800-1200 mg per day might be associated the long-term maintenance trial and began this with a continued risk of such side-effects. Further- within 48 hours of their last endoscopic examination. more, a case has been reported where long-term Nine patients received metiamide in a dose of continued administration of metiamide was associ- 400 mg taken at bedtime, and the other eight patients ated with therapeutic failure because of low plasma had an apparently identical placebo which they concentration of the drug (Blair et al., 1975). If, took in the same way. (The tablets were admini- therefore, a lower dose could be used once the ulcer stered on a double-blind basis.) Unmarked, was healed, it might be possible to maintain ulcer specially prepared, Rennie antacid tablets were issued healing while reducing the risk of side effects. to each patient for use in case of recurrence of ulcer The finding that duodenal ulcer patients secrete pain. There were two stomal ulcer patients in each more acid overnight than normal (Winkelstein, treatment group. The trial was designed to last 1935; James and Pickering, 1949; Levin et al., 1958), 12 months, and no other ulcer treatment was given and the ability of single oral doses of metiamide while a patient remained in the trial. No specific grossly to reduce or even abolish this secretion dietary advice was given, and no restriction placed (Milton-Thompson et al., 1974) led us to use a nightly on the consumption of tobacco or alcohol. dose only in this trial. ASSESSMENT OF PATIENTS Received for publication 19 November 1976 Each patient was seen at weekly intervals, an d 438 Gut: first published as 10.1136/gut.18.6.438 on 1 June 1977. Downloaded from

Nocturnal metiamide treatment in the management of healed duodenal ulceration 439 questioned about possible ulcer symptoms. New group each month (one month being four weeks), supplies of the drug were issued, and returned drugs and comparison made using Fisher's Exact Test. counted. Laboratory tests were undertaken. To investigate whether metiamide induced a more prolonged remission it was necessary to make allow- RELAPSE OF ULCER DISEASE ances for the fact that not all the patients had A symptomatic relapse was deemed to have occurred completed the same length of follow-up. This was when ulcer pain was present during two consecutive done by comparing the number of weeks that a weeks, sufficiently severe to require the taking of patient's ulcer had been in remission with that in antacids. relapse-for example, a patient who had been entered into the trial nine months before it ended but whose LABORATORY STUDIES ulcer had recurred at six months would have 24 A full blood count, platelet count, and differential weeks in remission and 12 weeks in relapse. The two white blood cell count were made each week. Plasma groups were then compared using a x2 test. electrolytes, urea and creatinine, and serum aspartate aminotransferase, lactic dehydrogenase, and alkaline Results phosphatase were measured each month. Two patients withdrew, fit and well, after six weeks ENDOSCOPIC ASSESSMENT and four months respectively. The first patient has A standard method was used throughout the study. not been included in the results, and the second only After sedation with 30 mg pentazocine intravenously, up to three months, when endoscopy confirmed a diazepam was given intravenously, in 2 5 ml aliquots sustained remission. The patients remaining for until ptosis or dysarthria occurred. An end-viewing analysis are shown in Table 2. instrument was used initially (Olympus GIF-D). Hyoscine-N-butyl-bromide, 40 mg, was given intravenously to facilitate visualisation of the duodenal Table 2 Patients studied ulcer when necessary. Ifa good view was not obtained Metiamide Placebo by this method, a side-viewing instrument was 47-4 47 passed (Olympus GF-B2). At all subsequent examina- Mean age (yr) 9 http://gut.bmj.com/ Age range 31-55 27-68 tions the same instrument was used by the same Mean length of history (yr) 10-1 12-8 investigator. Range 1-20 3-30 Endoscopy was performed at three-monthly The differences between the metiamide and placebo groups were not intervals while remission of disease was maintained, statistically significant (p > 0-5 in each case). and at any time in the event of a symptomatic relapse occurring. The findings were recorded in a

standard manner. The presence or absence of an After the first two patients had completed the on September 26, 2021 by guest. Protected copyright. ulcer or ulcers was noted. In all patients, who had trial, metiamide was withdrawn because of the not undergone previous surgery, the amount of occurrence of further cases of agranulocytosis in the endoscopically visualised 'duodenitis' was recorded USA in patients receiving full dosage. As patients using the same scoring system (Table 1) employed had entered the trial at different times, and had in the full dose trial. reached different stages between 31 and 48 weeks there was a different potential length of time in the CALCULATION OF RESULTS trial for each patient, at the time of stopping the The numbers of patients who had relapsed or trial. remained in remission was noted for each treatment MAINTENANCE OF ULCER HEALING Table 1 Scoring systems used to assess duodenitis at During the period of the trial, two patients on endoscopy. metiamide relapsed, at eight and 28 weeks, but six out of the eight on placebo had relapsed. This Duodenitis Score difference just failed to reach significance (p = 0-07). None 0 In Fig. 1 are illustrated the results in individual Severity In this figure is shown the period that each Slight 1 patients. Moderate 2 patient could have remained in the trial had the Severe 3 ulcer not recurred with the time when relapse Extent Localised small 1 occurred. More than one area 2 When the length of remission in each group was Throughout cap 3 compared with the period of relapse there was a Gut: first published as 10.1136/gut.18.6.438 on 1 June 1977. Downloaded from

440 M. H. Thompson, C. W. Venables, P. A. Smith, and W. Walker

I m Metiamide m Fig. 1 Ulcer remission in metiamide and placebo treated patients. The length of the bars represents the length of time the --.M patients could have been in the I trial. Time of ulcer relapse A A A A represented by start ofhatched 3 6 9 months 12 area. Vv Vv vv vv 0

I Placebo

highly significant prolongation of remission in the 8 patients treated with metiamide (Table 3). 6 4 REMISSION OF DUODENITIS Duodenitis The amount of duodenitis-that is, the combined score 2 http://gut.bmj.com/ score for the extent and severity of the change- 0 12 observed at repeated endoscopy is shown in Fig. 2. It can be seen that, in general, once the duodenitis 2 was reduced by 28 days' treatment using a full 4 therapeutic dose, this low-score situation was 6 maintained by nocturnal metiamide treatment. This 8 was not the case in the placebo group where a on September 26, 2021 by guest. Protected copyright. gradual return to the pretreatment state took place. Fig. 2 Remission of duodenitis. 0: healed ulcer at The scores for severity aloneshowed a similarpattern. endoscopy; 0: endoscopically-confirmed ulcer relapse. The figures 0-28 represent thefull-dose treatment SYMPTOMS period. All patients were symptom-free at the time of admission to the trial. Some symptomsofoesophageal reflux were noted occasionally during the trial, as pain was readily recognised as ulcer-pain by the was, rarely, mild epigastric pain. In general, though, patients, and again did not differ with the tablets the patients remained symptom free while in taken. On one occasion only, a defined symptomatic remission. There was no clear difference between relapse occurred and no ulcer was seen. The symptom metiamide-or placebo-treated patients while in continued, and the ulcer was visualised at endoscopy remission. When a symptomatic relapse occurred the one month later. On two occasions ulcer relapse was observed in asymptomatic patients, at regular endoscopy. Both patients had been taking placebo. Table 3 Number ofpatient-weeks ofrelapse and in each group. On the remaining 23 occasions that an endoscopic remission assessment was made, the symptomatic state and Total no. ofpatient weeks endoscopic findings were in agreement. In remission In relapse BIOCHEMICAL AND HAEMATOLOGICAL Metiamide 240 55 STUDIES Placebo 156 128 No haematological or biochemical abnormalities of x= 1335 r <0-001 marked degree occurred in any individual patient Gut: first published as 10.1136/gut.18.6.438 on 1 June 1977. Downloaded from Nocturnal metiamide treatment in the management of healed duodenal ulceration 441 which could be attributed to metiamide. There was, An important feature of this study was the use of however, a significant rise in the mean serum LDH repeated endoscopies to assess ulcer healing. This concentration from 129-7 units/i beforeanymetiamide proved to be remarkably acceptable and no patient was given to a peak of 195 7 units at three months refused a repeat examination. Although there was a (p = 0-002) (reference range 50-220), followed by a fairly good correlation between symptoms and gradual fall towards, but not reaching, pre-trial endoscopic findings, the fact that two ulcers occurred values. A similar rise and fall from 120-5 to 149-3 without any symptoms is ofconsiderable importance. units at four months occurred in the placebo group. This would suggest that repeat endoscopic assess- This change was not significant (p = 0O155). There ment is required in any study of ulcer relapse. was no accompanying change in serum AST or alkaline phosphatase concentrations. We thank Mr. B. W. Hawkins of Smith, Kline and There was a fall in the mean plasma potassium French Laboratories for supplies of metiamide and concentration from 4-3 mmol/l pre-trial to 3 9 comparable placebo tablets, and Messrs. Nicholas mmol/l after one month in the maintenance trial in Laboratories for supplies of unmarked Rennies the metiamide group. This change reverted to normal tablets. with continuing treatment. The mean plasma urea and creatinine concentra- References tions did not change significantly during the trial. Blair, E. L., Grund, E. R., Miller, I. T., Reed, J. D., Sanders D. J., Thompson, M. H., and Venables, C. W. (1975). Discussion Metiamide in the Zollinger-Ellison syndrome. American Journal of Digestive Diseases, 20, 1123-1130. Our earlier study (Thompson et al., 1975) had Burland, W. L., Duncan, W. A. M., Hesselbo, T., Mills, J. G., Sharpe, P. C., Haggie, S. J., and Wyllie, J. H. (1975). demonstrated that a single 28 day course of meti- Pharmacological evaluation of cimetidine, a new histamine amide was associated with healing of ulceration in H2-receptor antagonist, in healthy man. British Journal of most patients but it did not seem to induce a pro- Clinical Pharmacology, 2, 481-486. longed remission. This observation has been con- Carter, D. C., Forrest, J. A. H., Logan. R., Ansell, I., Lidgard, G., Heading, R. C., and Shearman, D. J. C. firmed in the present study by the behaviour of our (1975). Effect of the histamine H2 receptor antagonist 'placebo' treated patients and also appears to be a cimetidine on vagally induced gastric secretion in man. problem after treatment with the more recent H2 I.R.C.S. Journal of Medical Science, 3, 377. http://gut.bmj.com/ antagonist cimetidine as recently reported by Haggie Forrest, J. A. H., Shearman, D. J. C., Spence, R., and Celestin, L. R. (1975). Neutropenia Associated with et al. 1976. It could be argued that this does not metiamide. Lancet, 1, 392-393. matter as a further course of metiamide could be Haggie, S. J., Fermont, D. C., and Wyllie, J. H. (1976). given when the ulcer relapsed. However, an ulcer Treatment of duodenal ulcer with cimetidine. Lancet, 1, relapse can be associated with life threatening 983-984. Henn, R. M., Isenberg, J. I., Maxwell, V., and Sturdevant, complications, such as perforation or haemorrhage, R. A. L. (1975). Inhibition of gastric acid secretion by and may be asymptomatic. cimetidine in patients with duodenal ulcer. New England on September 26, 2021 by guest. Protected copyright. This study has shown that nocturnal therapy with Journal of Medicine, 293, 371-375. metiamide appears to reduce the number of relapses James, A. H., and Pickering, G. W. (1949). The role of gastric acidity in the pathogenesis of peptic ulcer. Clinical and significantly to increase the length of remission. Science, 8, 181-210. Before such a therapy could be widely adopted it Levin, E., Kirsner, J. B., Palmer, W. L., and Butler, C. would need to be shown to be without side-effects. (1948). Nocturnal gastric secretion. Archives of Surgery, There were no haematological complications in any 56, 345-356. Milton-Thompson, C. J., Williams, J. G., Jenkins, D. J. A., of our patients but there was a slight rise in serum and Misiewicz, J. J. (1974). Inhibition of nocturnal acid LDH concentrations. That a similar rise occurred in secretion in duodenal ulcer by one oral dose of metiamide. our placebo-treated patients suggests that a slight Lancet, 1, 693-694. laboratory variation accounts for this change. Sup- Thompson, M. H., Reed, J. D., Dale, G., and Venables, was no C. W. (1975). Early clinical experience with metiamide. port for this comes from the fact that there A histamine H2-receptor antagonist, in patients with associated change in AST concentrations. There were duodenal ulcer. American Journal of Digestive Diseases, no subjective side-effects and no patient complained 20, 1135-1141. of any difficulty in taking the treatment. Returned Lancet (1975). Treatment of duodenal ulcer by metiamide- treat- A multicentre trial. Lancet, 2, 779-781. drug counts suggest that patients took their Winkelstein, A. (1935). 169 studies in gastric secretion during ment regularly but this cannot be assured beyond all the night. American Journal of Digestive Diseases and doubt. Nutrition, 1, 778-782.