http://orcid.org/0000-0002-7439-575X
Figures
Abstract
Background and aims
Patients that have failed therapy for Helicobacter pylori (H. pylori) infection are incompletely characterized. The aim of this study was to characterize a H. pylori treatment resistant cohort compared to the cohorts of newly diagnosed, earlier eradicated and non-infected.
Material and methods
Patients were selected from routine referrals to the Endoscopy units at three different Norwegian hospitals. In all four cohorts, gastric biopsies were scored according to the Sydney classification, and symptoms according to the Gastrointestinal Symptom Rating Scale score, including sub-scores for upper gastrointestinal symptoms and functional bowel symptoms. Patients in the H. pylori resistant group were treated with a triple therapy regimen that consisted of levofloxacin, amoxicillin and a proton pump inhibitor.
Results
We included 185 patients, 42 H. pylori treatment resistant, 50 newly diagnosed, 61 previously H. pylori eradicated and 32 never infected. The treatment-resistant cohort had higher scores for upper gastrointestinal symptoms and functional bowel symptoms compared to the other groups except for the group being never H. pylori infected. The H. pylori resistant patients had lower Sydney scores than patients with newly diagnosed H. pylori infection. The triple combination showed a high efficacy of 91% to eradicate H. pylori.
Conclusions
Patients with treatment-resistant H. pylori infection had more gastrointestinal symptoms, but a lower Sydney score than patients with newly diagnosed infection. A treatment regimen including levofloxacin showed a high efficacy in eradicating H. pylori in patients that previously had failed eradication treatment.
Citation: Nestegard O, Johnsen K-M, Sørbye SW, Halvorsen F-A, Tønnessen T, Paulssen EJ, et al. (2020) Clinical characterization of Helicobacter pylori infected patients 15 years after unsuccessful eradication. PLoS ONE 15(9): e0238944. https://doi.org/10.1371/journal.pone.0238944
Editor: Iddya Karunasagar, Nitte University, INDIA
Received: April 21, 2020; Accepted: August 26, 2020; Published: September 23, 2020
Copyright: © 2020 Nestegard et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: The project is funded by Gastroenterological section, University hospital of Northern Norway, 9038 Tromsø, ref. prof. Jon Florholmen. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: No authors have competing interests
Introduction
Helicobacter pylori (H. pylori) has been identified as the main pathogenic factor for gastric and duodenal peptic ulcer.[1,2] Treatment of H. pylori infection cures most of the patients with ulcer, and the disease is no longer a chronically recurrent and disabling condition in Western world. However, in approximately 50% of the population worldwide, the infection is still a great problem, especially in underdeveloped countries.[3,4]
Initially, the most effective eradication regimes gave an efficacy >90%.[5] Unfortunately, during the 30 years of treatment one has observed a growing antimicrobial resistance of H. pylori[6], including resistance to clarithromycin and metronidazole, proposed to be a consequence of increased antibiotic use in the general population. Clarithromycin resistance has especially been of major negative impact as it was a main constituent of the recommended first line triple therapy. Metronidazole resistance is highly prevalent but is possible to overcome to some extent.[7] Thus, an increasing prevalence of resistance has been observed.[8,9] In areas with high occurrence of both clarithromycin and metronidazole resistance, a bismuth-containing quadruple therapy is recommended.[10] Despite this, a cohort of patients with treatment-resistant H. pylori has accumulated even in areas with low prevalence of antibiotic resistance, like Norway.[9] These patients will have a persistent gastric inflammation and a risk of recurrent ulcer disease and gastric cancer. There are few reports of the clinical or microbiological characteristics of this cohort, but some data has been presented in the search for a new effective antibiotic therapy, including levofloxacin.[11–13]
The goal of this study was to characterize a cohort with treatment-resistant H. pylori infection, and make clinical comparisons to patients with newly diagnosed H. pylori infection, previously successfully eradicated H. pylori, and a non-infected cohort. Finally, we wanted to evaluate the response to treatment with levofloxacin.
Materials and methods
Study population, enrolment and patient flow
Patients were recruited in the period 1990–2012 for screening in the Chronical Infection of Helicobacter Pylori (CRIHEP) study from three different gastrointestinal units at Norwegian hospitals: University Hospital of North Norway, Tromsø; Vestre Viken Hospital, Ringerike; and Vestre Viken Hospital, Drammen; Norway.
Four groups were invited to participate: The treatment-resistant H. pylori group (n = 42) were patients who had two or more unsuccessful treatment attempts for H. pylori infection. They were selected from medical records from 1990 to 2012 at the three different hospitals, and were invited to participate by a mail request. Responders were included after a new gastroscopy examination. The three additional patient groups were outpatients referred to the Endoscopy unit for gastroscopy: The newly diagnosed H. pylori infection group (n = 50) were included after positive findings of H. pylori infection at the endoscopic examination, the group of previously successfully treated H. pylori infection 5–20 years ago (n = 61), and the group of patients without any diagnosed H. pylori infection ever (n = 32) also underwent upper endoscopy. Patients with severe disease and those who were non-compliant were excluded.
Biopsies were obtained at endoscopy from the gastric corpus and pyloric antrum for histological examination and H. pylori rapid urease test.
Patients in the treatment-resistant H. pylori group were treated with oral omeprazole 20 mg b.i.d, amoxicillin 1 g b.i.d. and levofloxacin 500 mg b.i.d. for 10 days. Those with allergy to penicillin received metronidazole 500 mg t.i.d instead of amoxicillin.
Patients with previously untreated H. pylori infection and findings of peptic ulcer, erosive gastritis and/or erosive duodenitis were treated with oral omeprazole 20 mg b.i.d., amoxicillin 1 g b.i.d. and clarithromycin 500 mg b.i.d. for seven days. In patients with allergy to penicillin, metronidazole was used instead of amoxicillin as described above.
All subjects that were given treatment for H. pylori infection underwent a follow-up gastroscopy after 3–6 months upon which biopsies were obtained for the various examinations described above.
All patients gave an oral and written consent before enrolment in the study. The study was approved by the Regional Committee for Medical and Health Research Ethics (REC-North, ID:2009/2176-11), including approved the storage of biological material.
Symptom scores
All patients filled out the validated Gastrointestinal Symptoms Rating Scale (GSRS) questionnaire.[14] The GSRS includes 15 questions about symptoms where the answers are rated on a 7-graded scale, 1 representing absence of symptoms and 7 most severe symptoms.
To compare the gastrointestinal complaints among the four groups, comparisons were made between the various dimensions of the GSRS score, and yes/no according to reflux disease, dyspepsia and functional bowel symptoms according to their definitions.
For upper gastrointestinal symptoms, the questions were divided into 5 clinically relevant dimensions of the GSRS: abdominal pain, indigestion (increased flatus), reflux, diarrhoea and constipation, and the total score based on a study of patients referred to upper endoscopy for dyspepsia according to Dimenas et al. and Reviecki et al.[15,16] (Table 2)
The GSRS questionnaire defines ‘‘reflux syndrome” by asking about ‘‘heartburn” or ‘‘acid reflux”. The resulting ‘‘reflux syndrome” score is defined as the mean score of 2 or more of the two items heart burn and acid reflux in GSRS registration according to the references, and also to the Montreal definition from 2006 (Table 3).[14–17] The questions defining dyspepsia were the GSRS dimensions: “Have you had abdominal pain located in the upper abdomen for at least one week”, and “Have you ever had heartburn or acid regurgitation almost daily for at least one week”. A positive answer to at least one of the questions defined dyspepsia as used by Asfeldt et al.[18] (Table 3)
The dimensions most relevant for functional bowel symptoms (FBS) were the ten GSRS questions: abdominal pain, hunger discomfort, borborygmus, abdominal distension, flatulence, constipation, diarrhea, loose stool, urgency and incomplete evacuation. These are the GSRS dimensions that closest resemble irritable bowel disease and has been developed and validated by Wicklund et al.[19] A cut-off value of ≥22 defined a positive diagnosis of functional bowel symptoms based on best comparison (κ = 0.535) described in a previous study from our research group.[20] (Table 3).
Registrations and analyses
Clinical data including ongoing medication were registered. Biopsies were examined with the rapid urease test for on-site H. pylori diagnosis. Histological examination (hematoxylin-eosin staining) including detection was performed by one person (SWS) for H. pylori diagnosis and Sydney classification. A positive H. pylori status was based on a positive result for at least one of the two tests (H. pylori rapid -urease test and histological test). Moreover, a negative H. pylori status was indicated when both tests were negative.
Statistics
Statistical analyses were performed to evaluate difference between H. pylori groups in upper gastrointestinal symptoms defined by mean score of the five dimensions of GSRS questionnaire. For analyzing differences between two groups, pairwise Mann-Whitney tests were performed. As a global test for testing differences between groups a non-parametric method, Kruskal-Wallis one-way ANOVA was performed.
To detect potential confounders for reflux, dyspepsia and functional bowel syndrome symptoms, variables were tested in a logistic regression model using an Omnibus logistic regression model. The following variables were tested: H. pylori, sex, age, body mass index, snuff and smoking habits, and use of platelet inhibitors, NSAIDs, PPI, H2 blockers as well as H. pylori. Variables with p-values ≤0.10 were included in the multivariate models. Multivariate analyses were stratified for sex and the use of PPI and H2 blockers.
All statistical analyses was carried out in IBM SPSS Statistics 24 (IBM Corporation, Armonk, New York, USA).
Results
Patients included
A total of 185 patients, 165 in Tromsø, 18 in Drammen and 2 in Hønefoss, were included in the study. The baseline characteristics of the four study groups are shown in Table 1. Thirty-six of the 42 patients in the H. pylori resistant group, and 40 of the 50 patients in the newly H. pylori diagnosed group were treated according to the protocol. The remaining patients refused to participate in the treatment protocol.
Values are presented as number (ratio (%)) or mean (95% CI). For further details, see text.
Upper GI symptoms
The upper GI symptoms registered as total score of five dimensions (abdominal pain, indigestion, diarrhea, reflux and constipation [15,16]) in the five patient groups are shown in Table 2. The global test showed a p-value of 0.002 indicating a statistical significant difference in upper gastrointestinal symptom burden between the groups There was a significant difference in upper gastrointestinal symptom burden between groups. For the five dimensions of upper GI symptoms, the H. pylori resistant group showed a significant increased score compared to both the group of newly diagnostic H. pylori and H. pylori eradicated patients, but not to patients never infected with H. pylori (Table 2, Fig 1). In Table 2 each of the five symptom dimensions are also shown. No significant differences between the groups were observed regarding the prevalence of reflux syndrome and dyspepsia (Table 3).
P-values represent comparison to the H. pylori resistant group.
Functional bowel symptoms
Functional bowel symptoms (FBS) based on GSRS dimensions that closest resemble irritable bowel syndrome according to Wicklund et al [19] are shown in Table 3 and Fig 2. FBS was more frequent observed in the H. pylori resistant group than in the newly diagnosed H. pylori and the H. pylori eradicated patients, but not when compared to patients never H. pylori infected.
P-values represent comparison to the H. pylori resistant group.
Endoscopic characteristics
The endoscopic findings are shown in Table 4. Gastritis was more prevalent in H. pylori positive patient groups compared to the H. pylori treated group and the H. pylori patient groups never infected group. Moreover, no significant differences were observed between the two infected H. pylori positive groups. No patients in any group had duodenal ulcer.
Histological characteristics
In Table 5, the Sydney histological scores are presented. The treatment resistant H- pylori patients had a lower score than the newly diagnosed H. pylori patient group. Moreover, the group of H. pylori eradicated patients and never H. pylori patient groups infected had significantly lower score that the two other groups.
In Table 6 the Ki-67 index and presence of atrophy in gastric biopsies are evaluated. For both parameters no significant differences were observed between the two H. pylori patient groups infected groups including the subgroups before and after eradication (total of 4 subgroups). Moreover, when comparisons were performed between each of these 4 groups to the group of previously H. pylori eradicated and the group never infected significant differences were observed and especially for Ki-67 as shown in Table 6. Significant difference between the H. pylori eradicated and never infected group was also shown in Ki-67 (medium-high). To be noted, there was no increased atrophy in the H. pylori resistant group compared to the newly diagnosed H, pylori patient groups diagnosed group nor the H. pylori patient groups eradicated group. Finally, atrophy was slightly reduced after H. pylori patient groups eradication in the two H. pylori patient groups infected groups but not at significant levels.
H. pylori patient groups, diagnostics
The H. pylori diagnostics (at least one positive test of H. pylori-urease test and a positive histological test of H. pylori) in the four different H. pylori patient groups groups are shown in Table 7. Histological H. pylori patient groups detection in biopsies was positive in a few biopsies with negative urease test.
H. pylori patient groups, treatment
Thirty-six of the 42 patients in the H. pylori patient groups resistant group and 40 of the 50 patients in the newly H. pylori patient groups diagnosed group were treated according to the protocol. The Sydney score was significantly reduced in both treatment groups (Table 5)
Discussion
We have made comparisons between a cohort of patients with treatment- resistance H. pylori infection and patients who underwent gastroscopy due to upper gastrointestinal complaints and who were further sub-grouped into “newly diagnosed H. pylori patient groups infection”, “previously H. pylori patient groups eradicated” and “never H. pylori patient groups infected”. The treatment-resistant cohort had a higher upper gastrointestinal symptom score, more functional bowel symptoms (FBS), and a reduced histological inflammation (Sydney) score compared to the newly diagnosed H. pylori patient groups group. Levofloxacin in combination with amoxicillin and omeprazole was highly effective in the eradication of H. pylori in the resistant patients and even more effective compared to newly H. pylori diagnosed patients treated with a conventional eradication regime (amoxicillin, clarithromycin and omeprazole). As far as we know this is the first broad characterization of patients with resistant H. pylori infection. Finally, the two non-infected groups had low Sydney scores, but only the never infected group–also defined as “true” functional dyspepsia–scored high both for upper gastrointestinal symptoms and FBS. This may imply two different clinical phenotypes.
The upper gastrointestinal symptom score was higher in the H. pylori resistant group compared to the group of newly diagnosed H. pylori infection, and even more increased compared to the previously (>5 years) eradicated H. pylori group. This score is based on the baseline Gastrointestinal Symptoms Rating Scale (GSRS) of five dimensions of upper gastrointestinal symptoms: abdominal pain, indigestion, reflux, diarrhoea and constipation, and the total score based on a study of patients referred to upper endoscopy for dyspepsia as validated in previous reports.[15,16] As seen for reflux in Table 2 and for reflux syndrome in Table 3 no significant differences were observed between the groups with and without active H. pylori infection. The role of H. pylori in reflux is also somewhat controversial, as both the presence and the absence of an association have been reported.[21–23] In a population-based study from our research group, H. pylori was protective against reflux symptoms in men.[24] The functional bowel symptom score was also increased in the H. pylori resistant group compared to the group of newly diagnosed H. pylori infection and even more increased compared to the previously H. pylori eradicated group. The association between H. pylori infection and IBS is not settled. H. pylori infection has been shown to be a risk factor for IBS in one study, but not in three other reports.[25–28] There are, however, no larger, population-based studies that have addressed this issue except from that reported by Breckan et al. in 2012 where H. pylori infection was not found to be associated with functional bowel symptoms.[20]
We used a combination of levofloxacin and amoxicillin in the second line triple therapy of H. pylori. This antibiotic combination has been reported to be successful in second-line H. pylori eradication with an efficacy of some 90%, as seen in our study.[11,13] An even higher score has been shown in sequential or concomitant use of levofloxacin, with a cumulative therapeutic efficacy of as high as 97.8% in second line therapy.[10,12] Despite the highly efficacy of levofloxacin and amoxicillin in second line triple therapy of H. pylori in our and other studies, it should be noticed that antibiotic resistance against H. pylori is an increasing worldwide health problem that now also includes levofloxacin resistance.[29] Therefore, the choice of antibiotics should be carefully selected to prevent antimicrobial resistance, and should be based on antibiotic resistance testing, including testing for levofloxacin according to the 2017 Maastricht Guidelines.[10]
In our study, gastric atrophy was more pronounced in the treatment resistant H. pylori group compared to the group never infected with H. pylori. This is in agreement with other reports including a meta-analysis by Adamu et al.[30] Moreover, H. pylori eradication reduced gastric atrophy as shown, yet only slightly. This may be explained by a short observation time after eradication (3 months). A meta-analysis by Rokka showed that H. pylori eradication indeed reduced gastric atrophy,[31] which implies that eradication of H. pylori in apparent treatment resistant cases should be given more priority in order to reduce which a proposed gastric precancerous condition.[32]
Of interest was the cohort of patients never infected with H. pylori—a group defined as “true” functional dyspepsia according the Maastricht V/Florence Consensus Report from 2017.[10] As expected, these patients presented both upper gastrointestinal symptoms and functional bowel symptoms (Tables 1 and 2, Figs 1 and 2) on a level with the H. pylori resistant patients. Conversely, the previously H. pylori eradicated patients had low Sydney and endoscopic gastric scores comparable to the never H. pylori infected, but this formerly infected group had low upper gastrointestinal symptom and functional bowel symptom scores. This may indicate that H. pylori infection, when being treatment resistant and/or infection being longstanding, may precipitate functional abdominal disorders including dyspepsia. One could propose that a successful eradication of H. pylori infection would reduce such gastrointestinal complaints. However, this awaits further studies.
The strength of this study is that we have performed a broad characterization of a H. pylori treatment resistant group and made comparisons to relevant patient groups with upper gastrointestinal symptoms. Moreover, we have tested for potential confounders when comparing groups of patients. The weakness of the study is that a follow-up study (>5 years after H. pylori eradication) has to be performed to characterize potential functional gastrointestinal symptoms.
In conclusion, patients with treatment-resistant H. pylori infection had more upper gastrointestinal symptoms and functional bowel symptoms, whereas the Sydney score was lower than in patients with newly diagnosed H. pylori infection. A triple combination of levofloxacin, amoxicillin and omeprazole showed a high efficacy in eradicating the infection.
References
- 1. Warren J. R. & Marshall B. Unidentified curved bacilli on gastric epithelium in active chronic gastritis. Lancet (London, England) 1, 1273–1275 (1983).
- 2. Nomura A., Stemmermann G. N., Chyou P. H., Perez-Perez G. I. & Blaser M. J. Helicobacter pylori infection and the risk for duodenal and gastric ulceration. Annals of internal medicine 120, 977–981 (1994). pmid:7741826
- 3. Go M. F. Review article: natural history and epidemiology of Helicobacter pylori infection. Alimentary pharmacology & therapeutics 16 Suppl 1, 3–15 (2002).
- 4. Breckan R. K. et al. The All-Age Prevalence of Helicobacter pylori Infection and Potential Transmission Routes. A Population-Based Study. Helicobacter 21, 586–595, (2016). pmid:27172105
- 5. Lind T. et al. Eradication of Helicobacter pylori using one-week triple therapies combining omeprazole with two antimicrobials: the MACH I Study. Helicobacter 1, 138–144 (1996). pmid:9398894
- 6. Megraud F. & Lehours P. Helicobacter pylori detection and antimicrobial susceptibility testing. Clinical microbiology reviews 20, 280–322, (2007). pmid:17428887
- 7. Malfertheiner P. et al. Current concepts in the management of Helicobacter pylori infection: the Maastricht III Consensus Report. Gut 56, 772–781, (2007). pmid:17170018
- 8. Fischbach L. & Evans E. L. Meta-analysis: the effect of antibiotic resistance status on the efficacy of triple and quadruple first-line therapies for Helicobacter pylori. Alimentary pharmacology & therapeutics 26, 343–357, (2007). pmid:17635369
- 9. Larsen A. L., Ragnhildstveit E., Moayeri B., Eliassen L. & Melby K. K. Resistance rates of metronidazole and other antibacterials in Helicobacter pylori from previously untreated patients in Norway. APMIS: acta pathologica, microbiologica, et immunologica Scandinavica 121, 353–358, (2013). pmid:23083455
- 10. Malfertheiner P. et al. Management of Helicobacter pylori infection-the Maastricht V/Florence Consensus Report. Gut 66, 6–30, (2017). pmid:27707777
- 11. Di Caro S. et al. Levofloxacin/amoxicillin-based schemes vs quadruple therapy for Helicobacter pylori eradication in second-line. World journal of gastroenterology 18, 5669–5678, (2012). pmid:23155306
- 12. Gisbert J. P. et al. Second-line rescue triple therapy with levofloxacin after failure of non-bismuth quadruple "sequential" or "concomitant" treatment to eradicate H. pylori infection. Scandinavian journal of gastroenterology 48, 652–656, (2013). pmid:23556551
- 13. Katelaris P. H. & Katelaris A. L. A prospective evaluation of levofloxacin-based triple therapy for refractory Helicobacter pylori infection in Australia. Internal medicine journal 47, 761–766, (2017). pmid:28345276
- 14. Svedlund J., Sjodin I. & Dotevall G. GSRS—a clinical rating scale for gastrointestinal symptoms in patients with irritable bowel syndrome and peptic ulcer disease. Digestive diseases and sciences 33, 129–134 (1988). pmid:3123181
- 15. Dimenas E. et al. Quality of life in patients with upper gastrointestinal symptoms. An improved evaluation of treatment regimens? Scandinavian journal of gastroenterology 28, 681–687 (1993). pmid:8210982
- 16. Revicki D. A., Wood M., Wiklund I. & Crawley J. Reliability and validity of the Gastrointestinal Symptom Rating Scale in patients with gastroesophageal reflux disease. Quality of life research: an international journal of quality of life aspects of treatment, care and rehabilitation 7, 75–83 (1998).
- 17. Vakil N., van Zanten S. V., Kahrilas P., Dent J. & Jones R. The Montreal definition and classification of gastroesophageal reflux disease: a global evidence-based consensus. The American journal of gastroenterology 101, 1900–1920; quiz 1943, (2006). pmid:16928254
- 18. Asfeldt A. M. et al. Changes in the prevalence of dyspepsia and Helicobacter pylori infection after 17 years: the Sorreisa gastrointestinal disorder study. European journal of epidemiology 23, 625–633, (2008). pmid:18704703
- 19. Wiklund I. K. et al. An irritable bowel syndrome-specific symptom questionnaire: development and validation. Scandinavian journal of gastroenterology 38, 947–954 (2003). pmid:14531531
- 20. Breckan R. K., Asfeldt A. M., Straume B., Florholmen J. & Paulssen E. J. Prevalence, comorbidity, and risk factors for functional bowel symptoms: a population-based survey in Northern Norway. Scandinavian journal of gastroenterology 47, 1274–1282, (2012). pmid:23061445
- 21. Sharma P. & Vakil N. Review article: Helicobacter pylori and reflux disease. Alimentary pharmacology & therapeutics 17, 297–305 (2003).
- 22. Moayyedi P. & Talley N. J. Gastro-oesophageal reflux disease. Lancet (London, England) 367, 2086–2100, (2006).
- 23. Fox M. & Forgacs I. Gastro-oesophageal reflux disease. BMJ (Clinical research ed.) 332, 88–93, (2006). pmid:16410582
- 24. Breckan R. K. et al. The impact of body mass index and Helicobacter pylori infection on gastro-oesophageal reflux symptoms: a population-based study in Northern Norway. Scandinavian journal of gastroenterology 44, 1060–1066, (2009). pmid:19593688
- 25. Ford A. C., Forman D., Bailey A. G., Axon A. T. & Moayyedi P. Irritable bowel syndrome: a 10-yr natural history of symptoms and factors that influence consultation behavior. The American journal of gastroenterology 103, 1229–1239; quiz 1240, (2008). pmid:18371141
- 26. Agreus L., Engstrand L., Svardsudd K., Nyren O. & Tibblin G. Helicobacter pylori seropositivity among Swedish adults with and without abdominal symptoms. A population-based epidemiologic study. Scandinavian journal of gastroenterology 30, 752–757 (1995). pmid:7481542
- 27. Locke C. R. 3rd et al. Helicobacter pylori and dyspepsia: a population-based study of the organism and host. The American journal of gastroenterology 95, 1906–1913, (2000). pmid:10950034
- 28. Kawamura A. et al. Prevalence of irritable bowel syndrome and its relationship with Helicobacter pylori infection in a Japanese population. The American journal of gastroenterology 96, 1946, (2001). pmid:11419864
- 29. Saracino I. M. et al. High prevalence of primary antibiotic resistance in Helicobacter pylori isolates in Italy. Journal of gastrointestinal and liver diseases: JGLD 21, 363–365 (2012). pmid:23256118
- 30. Adamu M. A., Weck M. N., Gao L. & Brenner H. Incidence of chronic atrophic gastritis: systematic review and meta-analysis of follow-up studies. European journal of epidemiology 25, 439–448, (2010). pmid:20585973
- 31. Rokkas T., Pistiolas D., Sechopoulos P., Robotis I. & Margantinis G. The long-term impact of Helicobacter pylori eradication on gastric histology: a systematic review and meta-analysis. Helicobacter 12 Suppl 2, 32–38, (2007). pmid:17991174
- 32. de Vries A. C. et al. Gastric cancer risk in patients with premalignant gastric lesions: a nationwide cohort study in the Netherlands. Gastroenterology 134, 945–952, (2008). pmid:18395075