Functional dyspepsia (FD) is a common disease with a series of recurrent upper abdominal pain or discomfort symptoms. Such symptoms cannot be explained by the organ injury as endoscope or microscope observed inflammation. However, FD and chronic gastritis co-exists in some FD patients. In a recent FUTURE study conducted in Japan, only 13.2% patients in 2595 were confirmed as FD and the others were chronic gastritis with FD symptoms.1 Clinical characteristics of patients with chronic gastritis were similar to those with FD, except for shorter symptom duration.2 Whether the gastritis treatment can improve the outcome of FD remains controversial. It is a clinical challenge to the endoscopists to decide whether the mucosal abnormality should explain the dyspeptic symptoms and should therefore be classified as organic dyspepsia instead of FD. It is likely a matter of severity in the endoscopic view.
All currently used gastro-protecting agents own multiple function on mucus secretion, prostaglandin family and anti-oxidation, etc. But none of them shows superiority on certain gastritis or FD, which leads to the prescription often bases on the clinician's individual arbitrary selection. Traditional gastro-protect agents like sucralfate is still used clinically while the general effect is not satisfied. Even proton pump inhibitors (PPI) also have some gastro-protecting role,3 the long-term usage of them is still controversial in control of symptoms in dyspeptic patients.4
Gefarnate (geranyl farnesylacetate), is quite an old antiulcer drug since was introduced in 1970's.5–7 Its role has been proved in rat model as inhibiting the stress-induced reduction of mucosal prostacyclin and prostaglandin E28,9 and increasing the hexosamine level.10–12 Recently, Ohta et al13 further demonstrated that gefarnate prevents acute gastric mucosal lesion progression in C48/80-treated rats possibly by suppressing mucus depletion, neutrophil infiltration, and oxidative stress in the gastric mucosa. However, whether these functions are relevant in clinic is not clear. Up to date, there is no large prospective study on the clinical effect of gefarnate on chronic gastritis. Most previous studies were on ulcer or drug-induced injury and the sample size was not powerful enough.
Therefore, the current study was designed as a multicenter, randomized and controlled trial. The major aim is to evaluate the effective rate of gefarnate on endoscopic manifestation in erosive gastritis patients with FD-like symptoms. Moreover, the intra-mucosal prostaglandins and myeloperoxidase will be tested as indexes of gastric-protection. By analyzing the relationship between dyspeptic symptoms, Helicobacter pylori (H. pylori) status and erosion severity, we may decide which index is more important for the future gastritis trial.
The study was carried out as an open, randomized, positive drug parallel-controlled clinical trial performed in six centers during August 2009 to August 2010. The cohort size is expected to be 300 cases and each center should enroll 50 cases. The sample size was calculated upon such hypothesis that the effective rate of gefarnate and sucralfate is 80% and 65% (α=0.05, β=0.20, power=80%). Then 138 paired cases in each group are needed and the final sample size was set as 300 after considering dropout.
Before randomization, all subjects will be tested the H. pylori status by taken biopsy from gastric antrum to perform the bacteria staining. Patients were then randomized into two groups (Figure 1), GEF group receiving oral gefarnate 300 mg/d (100 mg, t.i.d.) (Hui Jia Qiang-G®, Seiko Eiyo Yakuhin Co. Ltd, Japan) or SUC group receiving sucralfate 3.0/d (1.0, t.i.d.) (Shu Ke Jie®, Shanghai Xudong Pham. Ltd, China) for six weeks. A randomization number generated by SAS program associated with either gefarnate or sucralfate was assigned to each patient. After the six-week therapy, the patients were asked to repeat an endoscopy to record the index of inflammation. Safety monitoring was also conducted during the whole follow-up.
The study has not been registered elsewhere but the whole protocol has been approved by ethic committee in the leading institute (CHEC2008-023). All subjects should sign on a written informed consent prior to be recruited in the study and got the free drugs.
Criteria for inclusion were age 18 to 65, diagnosed chronic non-atrophic gastritis with erosive lesions viewed under endoscopy within one week. Erosive lesions are defined as superficial or protrude erosion in the stomach. The patient should have at least one symptom of postprandial fullness, early satiation, epigastric pain or epigastric burning (according to the ROME III criteria for FD).14
Exclusion criteria include patients with malignancy diseases, peptic ulcer or severe heart or pulmonary disease. Patient who has been administrated with drugs that may affect evaluation (non-steroidal anti-inflammatory drug (NSAID), proton pump inhibitor, H2-antagonist, anti-acid regents and antibiotics, etc.) during two weeks before enrolled will be excluded. Other situation such as pregnancy and allergic habitus is not considered suitable for the study.
Modified Lanza scoring system
To obtain the trial and control drug's effect on gastric erosive lesion, Lanza score was used as an important and objective index in our study. Erosive gastritis was determined by endoscopy finding according to Sydney System15 and modified Endoscopic Chronic Gastritis Classification Consensus (2006, China).16 Endoscopic images for each patient before and after therapy were recorded and scored instantly after the examination. The lesions were graded as following: 0 (no erosion), 1 (one or two erosive lesions limited in one area as antrum, corpus or fundus), 2 (three to five lesions but in the same area), 3 (lesions involving two areas but less than ten) and 4 (extensive lesions or more than ten). Only one special endoscopist in each center was authorized the examination on enrolled patients, and they all received training before the trial started. The drug effect was classified as four grades: healed, the Lanza score become zero; significantly effective, score become one or decrease two or more; effective, score drop one; ineffective, no change or increase after therapy. The first three conditions comprise the final effective rate.
Gastric mucosal inflammation evaluation
One biopsy sample from erosion area will be fixed soon and delivered to one special pathologist for the inflammation evaluation. Specimens were treated with Hematoxylin & Eosin stains. Following the Sydney System,17 the histological findings can be divided by the activity and chronic inflammation, representing the neutrophils or monocytes infiltration. Both inflammations can be graded into four classes: 0 (normal), 1 (mild), 2 (moderate), or 3 (marked).
Symptom improvement evaluation
The symptom alteration of postprandial distress or epigastric pain syndrome was evaluated and recorded by the subjects, daily in the first week and once at the end of weeks 2, 4 and 6. Each symptom was graded as 0 (non-exist), 1 (mild), 2 (obvious, partially disturbing daily life) and 3 (severe, disturbing daily life and need drugs). The sum score of symptoms was then calculated to evaluate the drug effect on symptom release. The definition of “significantly effective” is all symptoms disappear or summed score decreases 4 or more. “Effective” is defined as the score decreases 2 or more and “ineffective” indicates for the rest. The effective ratio is obtained by the accumulated rate of “significantly effective” and “effective”.
Mucosal PGI2, PGE2 and MPO measuring
According to the previous experience,18 the content change of molecules is not obvious in the non-erosive area, so only one specimen from the gastric erosive area will be test for prostacyclin (PGI2), prostaglandin (PGE2) and myeloperoxidase (MPO). Only subjects in two centers (Changhai & Zhejiang First Hospital) were tested for the molecules. The content of above factors in gastric mucosa was tested by enzyme-linked immuno sorbent assay (ELISA) method, following the instruction of USCN LIFE Company (Missouri City, TX, USA). ELISA on each sample has been repeated for three times to make a mean value.
All data were collected to the Clinical Epidemiology Center in Changhai Hospital. Data were then computerized and analyzed with SPSS 18.0 software (IBM Corporation, NY, USA). The intent-to-treat set (ITT) consists of the randomized patients. The per-protocol (PP) analysis was performed by using data only from subjects characterized by the criteria: (i) the completion of whole treatment; (ii) the availability of results useful for the primary aim; and (iii) no any major protocol violations. Demographic characteristics of the two groups before treatment were compared using the Student's t-test and Fisher's exact test. The effective rate comparison was made by chi-squared test. Within each group, symptom scoring, endoscopic and histological grading was compared before and after therapy by using Wilcoxon signed rank test. For PG and MPO results, paired Wilcoxon signed ranks test was used to compare the alteration. Mann-Whitney U test was used when comparing the score between the two groups. The relationship between symptom and endoscopic score was obtained by Spearman rank correlation analysis. All statistical tests were two-sided, with a 5% level of significance.
Demographic characteristics of patients
Finally 244 cases of chronic non-atrophic gastritis were enrolled but only 235 cases can be put in the ITT set (9 missed the data form, Figure 2). Among the 235 cases (118 in gefarnate and 117 in sucralfate), only 170 cases had the integrated information which can be used for the PP analysis (102 in gefarnate and 68 in sucralfate). The H. pylori positive rate in gefarnate and sucralfate group is close (31.4% vs. 32.4%, PP; 27.1% vs. 18.8%, ITT, P >0.05). As shown in Table 1, there were no significant differences between gefarnate and sucralfate groups in age, sex, disease course, alcohol habit, etc (P >0.05). For the PP set, there was no significant difference between the mean follow-up time of two groups ((49.3±12.2) days vs. (50.3±11.5) days, P>0.05).
Effects of gefarnate on Lanza score
With the six-week continuous therapy, gefarnate provided a total effective rate of 83.3% (85/102, 95% CI 76.0%-90.6%) on Lanza score, which was statistically higher than sucralfate (69.1% (47/68, 95% CI 58.0%-80.2%), P=0.029, PP analysis, Figure 3A). The difference still exists when we use ITT analysis (72% vs. 40.1%, P <0.001). Both gefarnate and sucralfate could significantly decrease the endoscopic score as shown in Figure 4A, the Lanza score dropped from 2.69±0.83 to 1.32±1.15 after six week treatment with gefarnate (P <0.001).
Effects of gefarnate on dyspeptic symptoms
After one-week treatment, gefarnate or sucralfate could dramatically alleviate the symptoms as shown in Figure 4D. The mean summed score at baseline in gefarnate and sucralfate groups is 4.80±0.22 and 5.03±0.29 (P=0.428 by Mann-Whitney U test). At the end of weeks 1, 2, 4, 6, the mean symptom score became 2.51, 1.86, 1.29, 0.97 (gefarnate) and 3.10, 2.56, 1.84, 1.44 (sucralfate), respectively. Since the second week, the difference between the two groups became significant (P=0.017, 0.039, 0.048 by Mann-Whitney U test). Also by ITT analysis, the effective rate of gefarnate on symptom release was higher than sucralfate (67.0% vs. 39.3%, P <0.001, Figure 3B).
Effects of gefarnate on gastric inflammation
Figure 4B shows the mean active inflammation scores in histological examination of the gastric mucosa at baseline and the end of week 6. In both gefarnate and sucralfate groups, score of activity significantly decreased after treatment (dropped from 0.86±0.10 to 0.23±0.05 in gefarnate; and from 0.90±0.13 to 0.47±0.09 in sucralfate, P <0.001 by Wilcoxon Signed Rank test), and gefarnate showed more effective than sucralfate on active inflammation (36.4% vs. 23.1%, P <0.05, ITT, Figure 3C). For the chronic inflammation, gefarnated decrease the score from 2.19±0.07 to 1.46±0.07 (P <0.001, Figure 4C) and also showed a superior effective rate on monocyte infiltration of 57.7% than sucralfate (24.8% by ITT, P <0.001, Figure 3D).
Influence ofH. pylorion gastro-protecting effect
By the subgroup analysis, we further evaluated the drug's effect on both H. pylori positive and negative groups. Before treatment, H. pylori negative group showed a significantly lower activity or chronic inflammation score than H. pylori positive group (mean of 0.54 and 1.93 vs. 1.56 and 2.75 in gefarnate, P <0.001, Figure 5A and 5B). Gefarnate provided a better decreased trend than sucralfate both in active and chronic inflammation. However, even with gefarnate, the histological inflammation after therapy in H. pylori positive group was still severer than negative group (0.38 vs. 0.16 for activity, P=0.022; and 1.84 vs. 1.29 for chronic, P <0.001, Figure 5A and 5B). For endoscopic or symptom score, there was no difference between H. pylori positive and negative group with either treatment (Figure 5C and 5D).
Effects of gefarnate on mucosal PGI2, PGE2 and MPO concentration
Totally 80 paired samples were collected for the test of PGI2, PGE2 and MPO concentration, with 41 in gefarnate and 39 in sucralfate group. For PGI2 and PGE2, gefarnate increased both of the median (ng/g) from 45.3 and 14.0 to 64.6 and 58.3 (P <0.001, paired Wilcoxon signed rank test, Table 2). Meanwhile, no significant alteration of PGI2 and PGE2 after sucralfate treatment can be found (P >0.05). And the difference between gefarnate and sucralfate was only evident in PGE2 induction (P=0.032 by Mann-Whitney U test). Furthermore, depressed MPO level could only be observed in gefarnate treated group (median from 3.56 ng/g to 1.3 ng/g, P=0.015) and the difference between groups is statistically significant (P=0.007 by Mann-Whitney U test).
Only four cases of adverse effects occurred in 0.8% (1/118, diarrhea) of gefarnate and 2.6% (3/117, two constipations and one nausea) of sucralfate group. All the events were judged mild and soon recovered. The results demonstrated that gefarnate is safe when compared with a routine drug sucralfate (P=0.370).
Depending on few published studies about gefarnate in vivo, this clinical trial shows a reliable result of gefarnate on chronic gastritis with FD-like symptoms. Gefarnate can provide a satisfied inflammation attenuation of gastric mucosa with 6 weeks therapy, both macroscopically and microscopically. The advantage of gefarnate also exists on its rapid improvement on the gastritis-associated FD symptoms. Gefarnate shows better effect than sucralfate on symptom releasing including the Rome III criteria mentioned syndromes. Meanwhile, this effect has been proved by using sucralfate, a traditional gastro-protecting regent as a positive control. Indeed, sucralfate is not an ideal control for its poor compliance of patients and common side effect such as constipation. Placebo will be the option in the future studies because psychological effect is obvious in treatment of functional disease.
The mechanism by which gefarnate could protect gastric mucosa protection was also partially proved in this study. Gefarnate induces a significantly up-regulated PGI2 and PGE2 level and a down-regulated MPO level in the erosive lesion of antrum. MPO is released by the neutrophils and attends the oxidative stress protocol.19 So the decreased MPO level by gefarnate in this study may due to the less activity of inflammation. The concentration of MPO may not represent the oxidative status in the mucosa, testing the activity of MPO may be the better choice. Unlike other gastro-protecting agents, gefarnate's advantage relies on the stimulation of prostaglandin and hexosamine secretion. The role of mucin secretion by gefarnate even leads to its application in treatment of dry eye syndrome.20,21 Rebamipide can also promote prostaglandin and mucus creation, decrease the free-oxygen and inhibit inflammation.18,22,23 Teprenone has less effect on PG promotion but has strong activity on heat shock protein HSP 70 stimulation.24,25 Ecabet sodium owns a special role of inhibiting pepsin and H. pylori.26,27 Therefore, we cannot make a decision which gastro-protecting drug is the first choice for the treatment of chronic gastritis.
We did not enroll those patients with severe dyspeptic symptoms but without severe endoscopic gastritis. We deduce that the gastro-protecting agent will not provide a comparable symptom release rate in the gastritis non-associated FD. However, future studies are encouraged to compare the treatment effect between gastritis-associated and non-associated FD. The mechanism of FD symptoms remains unclear and sometimes it is not accompanying with the objective observation. Hu et al28 recently reported the role of ginger on gastric motility and symptoms of FD. They found ginger stimulated gastric emptying and antral contractions in patients with FD, but had no impact on gastrointestinal symptoms or gut peptides. Therefore, FD may be treated separately according to sub-group classification.
The activity of gastritis is often related to H. pylori infection, bile reflux or anti-inflammatory drugs. We also found H. pylori significantly affect the results of gastro-protecting drugs on the mucosa inflammation, but not influence the outcome of symptoms or endoscopic score. The drug may release the H. pylori-associated active or chronic inflammation but cannot reach the level as the negative subject's. H. pylori eradication is recommended as the first choice in H. pylori-associated gastritis.
Moreover, the prostaglandin stimulation by gefarnate did not lead to more diarrhea result in the observed patients, which indicates its safety. Though the current study is by far the largest clinical trial on the effect of gefarnate, there are obvious limitations on the study design, as is not blinded and placebo-controlled. A high dropout rate as 27% makes the result deflective. Turkkan et al29 found an inverse relation between severity of symptoms and level of H. pylori induced gastric inflammation or oxidative stress in patients with functional dyspepsia. Therefore, symptom scores should be used as major index in clinical trials on FD while endoscopic or inflammation score should be used in trials on gastritis.
In conclusion, gefarnate is proved an effective outcome on the mucosal erosive inflammation in functional dyspepsia patients with confirmed gastritis. The current study provides a better comprehension on the mechanism of gastro-protection agents.
1. Kinoshita Y, Chiba T. Therapeutic effects of famotidine on chronic symptomatic gastritis: subgroup analysis from FUTURE study. J Gastroenterol 2012; 47: 377-386.
2. Kinoshita Y, Chiba T. Characteristics of Japanese patients with chronic gastritis and comparison with functional dyspepsia defined by ROME III criteria: based on the large-scale survey, FUTURE study. Intern Med 2011; 50: 2269-2276.
3. Irani S, Krevsky B, Desipio J, Kim-Jaffe J, Maqbool S, Fisher RS. Rapid protection of the gastroduodenal mucosa against aspirin-induced damage by rabeprazole. Aliment Pharmacol Ther 2008; 27: 498-503.
4. Raghunath AS, Hungin AP, Mason J, Jackson W. Symptoms in patients on long-term proton pump inhibitors: prevalence and predictors. Aliment Pharmacol Ther 2009; 29: 431-439.
5. Langman MJS, Knapp DR, Wakley EJ. Treatment of chronic gastric ulcer with carbonoxolone and gefarnate: a comparative trial. Brit Med J 1973; 3: 84-86.
6. Smith PM, Sladen GE, Beck ER, Bennett PN, Lennard JE, Langman JS. A double blind trial of carbenoxolone and geranyl farnesylacetate in gastric ulcer. Scand J Gastroenterol 1975; 10: 753-755.
7. Truelove SC, Rocca M. Treatment of chronic gastric ulcer with gefarnate: a long-term controlled therapeutic trial. Curr Med Res Opin 1976; 4: 218-222.
8. Kobayashi K, Arakawa T, Nakamura H, Chono S, Yamada H, Satoh H, et al. Effect of gefarnate on endogenous prostacyclin, prostaglandin E2 and thromboxane in water-immersed rats. Tohoku J Exp Med 1983; 140: 181-185.
9. Arakawa T, Nakamura H, Chono S, Satoh H, Yamada H, Ono T, et al. Difference in mode of action of cimetidine and gefarnate on endogenous prostacyclin, prostaglandin E2 and thromboxane in rat gastric mucosa. Tohoku J Exp Med 1983; 140: 407-412.
10. Akhtar MS, Munir M. Evaluation of the gastric antiulcerogenic effects of Solanum nigrum, Brassica oleracea and Ocimum basilicum in rats. J Ethnopharmacol 1989; 27: 163-176.
11. Akhtar AH, Akhtar MS, Khan MA. Comparative effect of cimetidine and gefarnate on normal and ulcerated albino rats. Pak J Pharm Sci 1992; 5: 37-45.
12. Murakami M, Oketani K, Fujisaki H, Wakabayashi T, Ohgo T, Okabe S. Effects of the antiulcer drug geranylgeranylacetone on aspirin-induced gastric ulcers in rats. Jpn J Pharmacol 1982; 32: 299-306.
13. Ohta Y, Kobayashi T, Imai Y, Inui K, Yoshino J, Nakazawa S. Effect of gefarnate on acute gastric mucosal lesion progression in rats treated with compound 48/80, a mast cell degranulator, in comparison with that of teprenone. Biol Pharm Bull 2005; 28: 1424-1430.
14. Drossman D. The functional gastrointestinal disorders and the Rome III process. Gastroenterol 2006; 130: 1377-1390.
15. Sipponen P, Kekki M, Siurala M. The Sydney System: epidemiology and natural history of chronic gastritis. J Gastroenterol Hepatol 1991; 6: 244-251.
16. Fang J, Liu W, Shi Y, Ge Z, Xiao S. China consensus on chronic gastritis. Chin J Gastroenterol (Chin) 2006; 11: 674-682.
17. Rugge M, Genta RM. Staging and grading of chronic gastritis. Hum Pathol 2005; 36: 228-233.
18. Du Y, Li Z, Zhan X, Chen J, Gao J, Gong Y, et al. Anti-inflammatory effects of rebamipide according to Helicobacter pylori status in patients with chronic erosive gastritis: a randomized sucralfate-controlled multicenter trial in China: STARS study. Dig Dis Sci 2008; 53: 2886-2895.
19. Elseweidy M, Taha M, Younis N, Ibrahim K, Hamouda H, Eldosouky M, et al. Gastritis induced by Helicobacter pylori infection in experimental rats. Dig Dis Sci 2010; 55: 2770-2777.
20. Peral A, Domínguez-Godínez CO, Carracedo G, Pintor J. Therapeutic targets in dry eye syndrome. Drug News Perspect 2008; 21: 166-176.
21. Gipson IK, Hori Y, Argüeso P. Character of ocular surface mucins and their alteration in dry eye disease. Ocul Surf 2004; 2: 131-148.
22. Haruma K, Ito M. Review article: clinical significance of mucosal-protective agents: acid, inflammation, carcinogenesis and rebamipide. Aliment Pharmacol Ther 2003; 18: 153-159.
23. Talley NJ, Riff DS, Schwartz H, Marcuard SP. Double-blind placebo-controlled multicentre studies of rebamipide, a gastroprotective drug, in the treatment of functional dyspepsia with or without Helicobacter pylori
infection. Aliment Pharmacol Ther 2001; 15: 1603-1611.
24. Mizushima T. HSP-dependent protection against gastrointestinal diseases. Curr Pharm Des 2010; 16: 1190-1196.
25. Niwa Y, Nakamura M, Miyahara R, Ohmiya N, Watanabe O, Ando T, et al. Geranylgeranylacetone protects against diclofenac-induced gastric and small intestinal mucosal injuries in healthy subjects: a prospective randomized placebo-controlled double-blind cross-over study. Digestion 2009; 80: 260-266.
26. Yang Kim H, Hahm KB, Choi MG, Rew JS, Seol SY, Chun HJ, et al. Prospective multi-center trial for the efficacy of ecabet sodium on the relief of dyspepsia in Korean patients with chronic gastritis. J Clin Biochem Nutr 2007; 41: 160-168.
27. Koizumi W, Tanabe S, Nakatani K, Ishido K, Nishimura K, Azuma M, et al. Quadruple therapy with ecabet sodium, omeprazole, amoxicillin and metronidazole is effective for eradication of Helicobacter pylori after failure of first-line therapy (KDOG0201 Study). J Clin Pharm Ther 2010; 35: 303-307.
28. Hu ML, Rayner CK, Wu KL, Chuah SK, Tai WC, Chou YP, et al. Effect of ginger on gastric motility and symptoms of functional dyspepsia. World J Gastroenterol 2011; 17: 105-110.
29. Turkkan E, Uslan I, Acarturk G, Topak N, Kahraman A, Dilek FH, et al. Does Helicobacter pylori-induced inflammation of gastric mucosa determine the severity of symptoms in functional dyspepsia? J Gastroenterol 2009; 44: 66-70.