SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Authorship
  8. Acknowledgements
  9. References
  10. Supporting Information

Background

Pernicious anaemia (PA) has an increased risk for gastric cancer (GC). It is not established whether PA patients need to undergo endoscopic/histological follow-up.

Aim

To provide a systematic overview of the literature on PA and the development of gastric cancer, to estimate the gastric cancer incidence-rate.

Methods

According to PRISMA, we identified studies on PA patients reporting the incidence of gastric cancer. Quality of studies was evaluated using the Newcastle-Ottawa Quality Assessment Scale. Meta-analysis on annual gastric cancer incidence rates was performed.

Results

Twenty-seven studies met eligibility criteria. 7 studies were of high, 6 of medium, 10 of low and 4 of very low quality. Gastric cancer incidence-rates ranged from 0% to 0.2% per person-years in 7 American, from 0% to 0.5% in 2 Asiatic, from 0% to 1.2% in 11 Northern European studies and from 0% to 0.9% in 7 studies from other European countries. The incidence-rates of gastric cancer ranged from 0% to 1.2% per person-years in studies which used gastroscopy, from 0.1% to 0.9% in those based on International Classification of Disease. Heterogeneity between studies was not statistically significant at the 5% level (Chi-squared test = 17.9, = 0.08). The calculated pooled gastric cancer incidence-rate was 0.27% per person-years. Meta-analysis showed overall gastric cancer relative risk in PA as 6.8 (95% CI: 2.6–18.1).

Conclusions

This systematic review shows a pooled gastric cancer incidence-rate in pernicious anaemia of 0.27% per person-years and an estimated nearly sevenfold relative risk of gastric cancer in pernicious anaemia patients. Further high quality studies are needed to confirm this higher risk.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Authorship
  8. Acknowledgements
  9. References
  10. Supporting Information

Pernicious anaemia (PA), also known as Biermer's disease, is considered an autoimmune disorder characterised by atrophic damage of the gastric body mucosa and, as a consequence, by the loss of parietal cells which normally produce intrinsic factor, a protein that firmly binds dietary vitamin B12 and favours its transport through the terminal ileum mucosa wall.[1] The deficit of intrinsic factor is the pathophysiological mechanism which leads to the deficiency of vitamin B12 and to the development of macrocytic anaemia, which is the more frequent clinical manifestation of PA. From the clinical point of view, PA can be insidious and have heterogeneous presentation. In general, symptoms appear slowly and can be related to anaemia or to neurological involvement with peripheral neuropathy or more severe complications.[2] Moreover, patients with PA can develop long-term complications, such as the development of gastric cancer (GC).

This risk of developing GC in patients with PA is reported in the literature, although the studies differ from each other in terms of sample size and methods for the identification of GC used during the follow-up period. Corpus-restricted atrophic gastritis with antrum-spared, which is typically seen in patients with PA, is not considered to be part of the precancerous cascade described by Correa.[3] In contrast, more extensive atrophy of the gastric mucosa (multifocal atrophic gastritis) seems to be associated with an increased risk of progression to gastric neoplastic lesions.[4, 5] Furthermore, the clinical onset of PA has been described as a possible last step in the natural history of atrophic gastritis which may have been present for many years.[6] In light of the above, different clinical management of patients with PA or with atrophic gastritis does not seem to be justified. Recently, a board of experts published recommendations about the management of precancerous conditions and lesions of the stomach, specifying that guidelines were not addressed to precancerous conditions such as PA.[7] At the same time, guidelines about management of GC proposed by the British Society of Gastroenterology and Surgical Oncology stated that, being a precancerous condition, PA may be considered for endoscopic monitoring, although precise indications are lacking.[8] Thus, the management of patients with PA remains controversial. The entity of the risk of developing GC in these patients still remains undetermined.

We aimed to review the literature available on PA and the development of GC to estimate the gastric cancer incidence-rate.

Materials and Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Authorship
  8. Acknowledgements
  9. References
  10. Supporting Information

Literature search strategy

The present review was based on published results of studies on PA which were identified via a systematic computerised search in PubMed. The database was searched for published articles up to September 15th, 2011. The full search strategy for PubMed is included in Appendix (Table S1), and involved the following search components: pernicious anaemia, atrophic gastritis, Type A gastritis, macrocytic anaemia, cobalamin deficiency, vitamin B12 deficiency, intrinsic factor deficiency, autoimmune gastritis, stomach cancer, gastric cancer, gastric adenocarcinoma, gastric neoplas*, gastric carcinoma, gastric tumour. The search was conducted according to PRISMA guidelines.[9] We also searched in reference lists of included studies.

Study selection

Inclusion and exclusion criteria

Studies were included in the systematic review if they fulfilled all of the following criteria: (i) observational study including patients with PA and reporting the numbers of GC identified during a defined follow-up period, (ii) the study was performed in adult patients (iii) the study was an original full paper which presented unique data. Studies were included without any time limit. Studies were excluded if (i) they were reviews, letters, editorials, case-reports, (ii) follow-up data were not available, (iii) language was other than English, Spanish, Italian, German or French.

Selection

Study selection was performed independently by two reviewers (LV and EL) on the basis of title and abstract. In case of doubt, the whole publication was evaluated. Differences were clarified by discussion with a third investigator (BA).

Data extraction

Two reviewers (LV and EL) independently extracted the following information from each publication: the first author's last name, year of publication, country of study location, study design, criteria for diagnosis of PA, sources of selection of participants, numbers of patients investigated, duration of follow-up period (years), calculation of person-years, age of patients (median or mean or range), gender, type of follow-up protocol (active or not), methods of follow-up, methods for identification of GC, histology, numbers of GC cases.

Quality assessment

Two reviewers (LV and EL) evaluated the quality of all included studies using the Newcastle-Ottawa Quality Assessment Scale for cohort studies.[10] This scale awards a maximum of nine stars to each study. Three categories are considered: Selection of studied cohort which includes four items, Outcome Assessment with three items and Comparability between studied cohort and controls with two items. A study can be awarded a maximum of one star for each numbered item within the Selection and Outcome Assessment categories and two stars can be given for Comparability. We arbitrarily defined studies of high quality those which obtained nine stars, medium quality those which scored seven or eight stars, low quality those which scored five or six stars and very low quality those studies which obtained four stars or less. Discrepancy in quality assessment was discussed and resolved by two reviewers (LV and EL).

Outcome

The outcome measure of interest was the incidence-rate of GC, which was calculated as the ratio between numbers of new GC identified during the follow-up period and person-years.

Statistical analysis

We extracted from each included study the number of incident GC cases and the number of person-years of exposure to risk. The incidence-rates per person-years were calculated by the reviewers if not explicitly stated. Statistical heterogeneity was assessed using Chi-squared test. Mann–Whitney test was used when appropriate. A P-value of <0.05 was considered statistically significant. The pooled rate and its 95% confidence intervals of incidence-rates of GC per person-years was calculated. To estimate the overall GC relative risk (RR) in PA patients, the annual incidence rates of GC in PA patients were calculated for the selected studies, in which GC diagnosis was made by an active endoscopic-histological follow-up. These GC incidence rates were then used in a meta-analysis, and compared with the national annual GC incidence rates of both genders in the general population aged over 40 years as reported for the single European countries by the website GLOBOCAN,[11] to calculate the risk of developing GC in patients with PA, assuming a constant annual incidence ratio. The statistical analysis was performed using a dedicated software package (version 11.3; MedCalc Software, Mariakerke, Belgium).

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Authorship
  8. Acknowledgements
  9. References
  10. Supporting Information

Search results

The electronic search strategy identified 15 387 records from Pubmed, 2760 of which were unique (Figure 1). These articles were screened on the basis of title and abstract and, after the application of our inclusion and exclusion criteria, 71 articles were retrieved for full-paper evaluation. Of these 71 full-papers, 27 met eligibility criteria and were subjected to data extraction.[12-38]

image

Figure 1. Flow-chart of study selection.

Download figure to PowerPoint

Quality assessment

Global quality assessment of included studies is shown in Table S2. Seven studies were deemed to be of high, six of medium, ten of low and, finally, four of very low quality. Of the six high-quality studies, four [19, 25, 28, 29] reported large sample sizes, representative of communities which were American and North European. The majority of the studies which were considered of low and very low quality were not matched cohort studies and the Comparability category of the Newcastle-Ottawa Quality Scale was not assessable.[12, 13, 16-18, 20-24, 31, 34, 37, 38]

Characteristics of included studies

The characteristics of the included studies are summarised in Table S3. Female gender was prevalent in 14 studies with a range between 55.3% and 81.2%, about 50% in four studies, only male patients were investigated in two studies, while in the remaining studies gender was not reported. Patients had a median age over 50 years in all studies. In particular, five studies included patients between 50 and 60 years,[27, 28, 30, 34, 36] six studies between 61 and 70 years[12, 13, 18, 22, 23, 33] eight studies over 71 years.[19, 20, 24-26, 29, 32, 38] The majority of the 27 studies were European, in particular 11 (40.8%) were from Northern Europe (Denmark, Finland, Sweden) and 7 (25.9%) from other European countries (United Kingdom, Italy, Spain, Germany). Seven studies (25.9%) were American and only two studies (7.4%) were performed in Eastern Countries (China). The design was prospective in 17 (63%) studies. Diagnostic criteria for the diagnosis of PA were carefully described in 22 (81.5%) studies. In particular, clinical criteria such as blood tests (macrocytic anaemia, low vitamin B12, iron, gastrin, pepsinogen I), positive Schilling test, gastric pH, megaloblastic bone marrow or neuropathy (subacute combined cord degeneration, peripheral neuropathy, dementia) were used in 17 studies[12, 13, 17, 18, 20, 24, 26, 27, 30-38]; data were extracted on the basis of International Classification of Disease (ICD) in four studies[19, 25, 28, 29] and using hospital records in one study.[14] Five studies [15, 16, 21-23] included patients with PA without providing further details. The sources of the patients differed: 14 studies included patients hospitalised for PA and after followed-up as out-patients,[12, 14, 20-24, 26, 27, 30-32, 35, 37] 4 studies included hospitalised patients representative of general population,[19, 25, 28, 29] 2 studies included both inpatients and out-patients [34, 38] and 3 studies included out-patients referred to haematological or gastroenterological units for anaemia or gastrointestinal symptoms.[18, 33, 36] Three studies [13, 16, 17] lacked information about the origin of participants, while in the Blackburn study,[15] patients were recorded in the files of eight PA clinics, but it was not indicated whether the patients were hospitalised or not. The studies were very variable with regard to the follow-up protocol. Only nine studies (33.3%) had an active follow-up based on different diagnostic tools: EGD with biopsies (n = 6),[17, 18, 23, 27, 34, 36] cytology (n = 1),[16] EGD or Xray (n = 1),[24] clinical/EGD (n = 1)[20] at regular intervals. In the majority of cases, patients were not actively followed-up (n = 18), but in three of these studies, diagnostic investigations were performed on the basis of new clinical symptoms.[12, 31, 38]

Identification of GC

Nine studies established the diagnosis of GC by EGD with biopsies,[17, 18, 20, 23, 27, 33, 34, 36, 38] six studies retrieved this information from death certificates of hospital registers,[14, 15, 21, 26, 30, 37] four studies used ICD on computer-based files,[19, 25, 28, 29] while the remaining studies made use of different methods: X-ray and surgery/autopsy (n = 3),[16, 31, 32] EGD or X-ray (n = 4),[13, 22, 24, 35] EGD/clinical interview/death certificates (n = 1).[12] Considering the incidence-rates of GC per person-years according to the countries where studies were carried out, we found that the GC incidence-rates ranged from 0% to 0.2% in 7 American studies, from 0% to 0.5% in 2 Asiatic studies, from 0% to 1.2% in 11 Northern European studies and from 0% to 0.9% in 7 studies of other European countries. The incidence-rates of GC per person-years ranged from 0% to 1.2% in studies which used EGD for identification of GC, from 0.2% to 0.9% in those which retrieved data from death certificates, from 0.1% to 0.4% in those which used ICD and from 0% to 0.6% in those which combined different methods. The GC incidence-rates per person-years were similar considering studies with only inpatients or out-patients and studies with retrospective or prospective design.

Heterogeneity between studies was not statistically significant at the 5% level (Chi-squared test = 17.9, degree of freedom = 11, P = 0.08). Globally, a total of 22 417 patients with PA were studied, for a total of 157 319.8 person-years with 417 new cases of GC. This figure corresponds to a pooled GC incidence-rate per person-year of 0.26%. As shown in Table 1, when we considered only the six studies in which an active follow-up protocol by gastroscopy and histology was performed, we observed a pooled GC incidence-rate per person-years of 0.27 (95% CI: 0.25–0.29) corresponding to 7 new GC cases observed in 2563.4 person-years.

Table 1. Characteristics of the studies in which an active follow-up by gastroscopy and histology was performed in PA patients
AuthorYearCountryPopulation (n)Person-YearsCases of GCIncidence of GC (CI 95%) (%)
Borch[17]1986Sweden61164.700
Bresky[18]2003Spain68≈54400
Elsborg[23]1977Denmark6826310.38 (0.34–0.46)
Kokkola[27]1998Finland7186920.23 (0.20–0.26)
Sjoblom[34]1993Finland5616821.19 (−0.77–3.15)
Vannella[36]2010Italy129554.720.36 (0.32–0.40)
Pooled rate  4532563.470.27 (0.25–0.29)

As shown in Figure 2, a meta-analysis performed by using the calculated annual GC incidence rates of the six studies shown in Table 1, showed an overall GC relative risk in PA as 6.8 (95% CI: 2.6–18.1) by considering random effects given the significance level of heterogeneity (Q 81.3; P < 0.0001).

image

Figure 2. Forest plot: annual gastric cancer incidence rates of the six studies with active endoscopic-histological follow-up in patients with pernicious anaemia.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Authorship
  8. Acknowledgements
  9. References
  10. Supporting Information

Gastric cancer is the most severe long-standing complication of PA and, to our knowledge, this is the first systematic review on the estimate of GC incidence-rate. We conducted a comprehensive search strategy including a great number of terms and their combinations without time limit and including the main languages. From 15 387 articles, we selected, according with our inclusion criteria, 27 papers published between 1950 and 2011 concerning a total of 22 417 patients with PA. In detail, 417 new cases of GC were identified in 157 319.8 person-years of follow-up, corresponding to an overall pooled GC incidence-rate per person-years of 0.26% in patients with PA with a range between 0% and 1.2%. To evaluate the quality of included studies, we applied the Newcastle-Ottawa Quality Scale for cohort studies, which permits a critical appraisal of data.[10] Unfortunately, we found that almost half of the included studies had low or very low quality. Although the 27 included studies showed similar incidence rates for GC, clinical and methodological heterogeneity was present. Europe and USA are well represented among the studies included in this review. On the contrary, the conclusions for Asiatic countries were based on only two Chinese studies.[20, 38] In fact, PA can affect all racial and ethnic groups and, although China is the most populous region country of the world, relatively little information is available.[39] The included studies are the two largest published in China, but studies with more precise diagnostic strategies and more rigorous methodology are needed to establish the true incidence of PA in Eastern countries.

The sources of patients were frequently hospitals, and the design was retrospective in 10 (37%) studies. However, we found that the incidence-rates of GC are similar among the studies irrespective of the study design and the source of patients.

An important difference among included studies is the methods used for follow-up. Only nine studies had an active follow-up,[16-18, 20, 23, 24, 27, 34, 36] only six of them used gastroscopy together with histological evaluation of biopsies.[17, 18, 23, 27, 34, 36] The different approach across the studies should take into account the decade in which the studies were performed. Although we identified old studies which used cytology[16] or gastrocamera[22] as diagnostic tools, these methods were completely replaced by X-ray imaging, which has been widely used in clinical practice during the 20th century. The introduction of the fiberscope and, subsequently, of the video endoscope, which permits direct vision of the upper GI tract, has been a revolution in the diagnosis of diseases of the stomach. The studies with active follow-up began in the 1980s when these tools became easier to use and more widely available in clinical practice. Regular follow-up by X-ray or EGD, irrespective of clinical symptoms, can theoretically lead to an earlier diagnosis of GC. Older studies based the incidence-rates of GC exclusively on death certificates, while those with an active follow-up used diagnostic tools, such as EGD or X-ray, for the early identification of GC, and this could make studies difficult to compare. These different methodological approaches to establish the diagnosis of GC (death certificates or X-ray) may have misclassified gastric tumours in patients with PA as carcinoma rather than gastric neuroendocrine tumours which are also common in these patients.[2, 17] In fact, the histological diagnosis was not systematically obtained in all studies, not being reported in nine studies.[14, 15, 19, 21, 28, 29, 35, 37] Considering this limitation, the rate of gastric adenocarcinoma could be lower of that reported. However, the pooled GC incidence-rate per person-year of 0.27% (7/2563) calculated on the six selected studies with active follow-up by gastroscopy and histology again overlapped the overall estimate obtained by considering all studies, supporting the robustness of our analysis.

Thus, the main finding of this systematic-review is the similarity of the incidence-rates of GC which are remarkably stable among all studies despite major heterogeneity between individual studies in terms of their design (country of study location, sources of selection of participants, sample size, type and methods of follow-up, methods for identification of GC).

It is difficult to compare this result with the incidence-rate of GC in the general population, mainly for two reasons: (i) the reference rates for the general population vary enormously according to the locality of the study, between different countries and even within different region of the same country, and it is not obvious what ‘general population’ should be used for the comparison, (ii) in almost all countries of the world the incidence of GC is decreasing steadily year by year; in our study we have made no time restriction and included studies from 1950 to 2011, making it very difficult to find reference rates of the general population.[40]

Nevertheless, within the context on these limits, we performed a meta-analysis based on the calculated annual GC incidence rates in PA patients of the selected six studies which performed active endoscopic-histological follow-up, and compared annual incidence rates of the general population aged over 40 years of age for the single countries in which the selected studies have been conducted. By this analysis, we obtained a 6.8-fold RR of GC in patients with PA. Although we are aware that the results of this meta-analysis may be biased by the limits explained above, we think, however, that they should be acceptable as an estimate of the overall probability of PA patients to develop GC and may represent a starting point to design further population-based prospective studies.

The range of annual incidence-rates for GC in PA patients overlaps that reported in literature for patients with atrophic gastritis which is between 0% and 1.8% per year.[41] This overlap is not surprising because it is known that PA is only one of the possible clinical presentations of atrophic gastritis. As shown in previous studies, patients with atrophic gastritis may present iron deficiency anaemia as well as PA.[6, 42, 43] In general, patients presenting with iron deficiency anaemia are younger and more frequently females compared with patients presenting with PA.[6, 44] The median age was over 50 in this systematic-review which confirms that PA is a condition which involves more typically the elderly. This may be explained by the fact that depletion of vitamin B12 may take many years and appears in the latest phase of the evolution of atrophic gastritis. Thus, although PA is typically associated to corpus-restricted atrophy, it may be a late clinical presentation of a well-known precancerous condition, such as multifocal atrophic gastritis. Unfortunately, we could not evaluate the type of gastritis underlying the diagnosis of PA because these data were not systematically described in the included studies. For the same reasons, another limitation of this systematic review is that it was not possible to take account of a known risk factor for GC, Helicobacter pylori (H. pylori). Indeed, almost fifty per cent of the included studies precede the discovery of H. pylori. Some studies reported that H. pylori infection is an infrequent finding in PA patients suggesting that the pathological process in PA protects the stomach from colonisation by H. pylori.[44, 45] Other studies showed that PA patients had high prevalence of H. pylori antibodies with low H. pylori staining in gastric mucosa reflecting a previous or hidden H. pylori infection.[46, 47] However, these topics are beyond the aim of this systematic review.

Recent guidelines proposed the standardisation of the management of patients with precancerous conditions and lesions such as atrophic gastritis, intestinal metaplasia or dysplasia to direct resources to patients with the greatest risk of GC, with a cost-effective program of surveillance.[7] Patients with atrophic gastritis and those with PA have similar clinical presentation (anaemia, deficit of vitamin B12, neurological involvement), similar histological lesions (atrophy, intestinal metaplasia) and similar incidence-rate for GC, and PA might be considered a similar disorder at a different point of the wide spectrum of clinical manifestations of atrophic gastritis.[2, 6] With regard to the need of endoscopic surveillance of PA, further studies are needed. Moreover, the increased risk for the developing of gastric neuroendocrine tumour in patients with PA could represent an additional potential rationale for endoscopic surveillance in these patients.

In conclusion, irrespective of heterogeneity between included studies, this systematic review shows a pooled GC incidence-rate in patients with PA of 0.27% per person-years and an estimated nearly sevenfold RR of GC in PA patients. Further high quality studies are needed to confirm this higher risk.

Authorship

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Authorship
  8. Acknowledgements
  9. References
  10. Supporting Information

Guarantor of the article: Bruno Annibale.

Author contributions: Vannella L. contributed to the selection of the studies, to the extraction of data and wrote the article. Lahner E. contributed to the selection of the studies and to the extraction of data. Osborn J. contributed to revision of statistical analysis and critically revised the manuscript. Annibale B. contributed to conception and design of the study and to final revision of the manuscript. All authors approved the final draft submitted.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Authorship
  8. Acknowledgements
  9. References
  10. Supporting Information

We thank Michele Putignano and Tiziana Mattei, the librarians of University Sapienza of Rome, for providing support for the literature search.

Declaration of personal interests: None. Declaration of funding interests: This study was supported by grants from the Italian Ministry for University and Research, PRIN 2007 and from University Sapienza of Rome, 2010–2011 Italy.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Authorship
  8. Acknowledgements
  9. References
  10. Supporting Information
  • 1
    Wintrobe MM, Lee GR, Boggs DR, et al. Megaloblastic and non megaloblastic macrocytic anaemias. In: Wintrobe MM, Lee GR, Boggs DR, Bithell TC, Foerster J, Athens JW, Lukens JN, eds. Clinical Haematology, 8th edn. Philadelphia: Lea & Febiger, 1981; 559604.
  • 2
    Lahner E, Annibale B. Pernicious anaemia: new insights from a gastroenterological point of view. World J Gastroenterol 2009; 15: 51218.
  • 3
    Correa P, Piazuelo MB. The gastric precancerous cascade. J Dig Dis 2012; 13: 29.
  • 4
    Miehlke S, Hackelsberger A, Meining A, et al. Severe expression of corpus gastritis is characteristic in gastric cancer patients infected with Helicobacter pylori. Br J Cancer 1998; 78: 2636.
  • 5
    Cassaro M, Rugge M, Gutierrez O, et al. Topographic patterns of intestinal metaplasia and gastric cancer. Am J Gastroenterol 2000; 95: 14318.
    Direct Link:
  • 6
    Hershko C, Ronson A, Souroujon M, et al. Variable hematologic presentation of autoimmune gastritis: age-related progression from iron deficiency to cobalamin depletion. Blood 2006; 107: 16739.
  • 7
    Dinis-Ribeiro M, Areia M, de Vries AC, et al. Management of precancerous conditions and lesions in the stomach (MAPS): guideline from the European Society of Gastrointestinal Endoscopy (ESGE), European Helicobacter Study Group (EHSG), European Society of Pathology (ESP), and the Sociedade Portuguesa de Endoscopia Digestiva (SPED). Endoscopy 2012; 44: 7494.
  • 8
    Allum WH, Blazeby JM, Griffin SM, et al. Association of upper gastrointestinal surgeons of Great Britain and Ireland, the British Society of Gastroenterology and the British Association of Surgical Oncology. Guidelines for the management of oesophageal and gastric cancer. Gut 2011; 60: 144972.
  • 9
    Moher D, Liberati A, Tetzlaff J; PRISMA Group, et al. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med 2009; 6: e1000097.
  • 10
    Wells GA, Shea B; O'Connell D, et al. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. Available at: http://www.ohri.ca/programs/clinical_epidemiology/oxford.htm. Accessed November 30, 2012.
  • 11
    GLOBOCAN 2008. International Agency for Research on Cancer. World Health Organization. Estimated cancer Incidence, Mortality, Prevalence and Disability-adjusted life years (DALYs) Worldwide in 2008. Available at: http://globocan.iarc.fr. Accessed November 30, 2012.
  • 12
    Armbrecht U, Stockbrügger RW, Rode J, et al. Development of gastric dysplasia in pernicious anaemia: a clinical and endoscopic follow up study of 80 patients. Gut 1990; 31: 11059.
  • 13
    Arvanitakis C, Holmes FF, Hearne E 3rd. A possible association of pernicious anemia with neoplasia. Oncology 1979; 36: 1279.
  • 14
    Berkson J, Butt HR, Comfort MW. Occurrence of gastric cancer in persons with achlorhydria and with pernicious anemia. Proc Staff Meet Mayo Clin 1956; 31: 58396.
  • 15
    Blackburn EK, Callender ST, Dacie JV, et al. Possible association between pernicious anaemia and leukaemia: a prospective study of 1625 patients with a note on the very high incidence of stomach cancer. Int J Cancer 1968; 3: 16370.
  • 16
    Boon TH, Schade RO, Middleton GD, et al. An attempt at presymptomatic diagnosis of gastric carcinoma in pernicious anemia. Gut 1964; 5: 26970.
  • 17
    Borch K. Epidemiologic, clinicopathologic, and economic aspects of gastroscopic screening of patients with pernicious anemia. Scand J Gastroenterol 1986; 21: 2130.
  • 18
    Bresky G, Mata A, Llach J, et al. Endoscopic findings in a biennial follow-up program in patients with pernicious anemia. Hepatogastroenterology 2003; 50: 22646.
  • 19
    Brinton LA, Gridley G, Hrubec Z, et al. Cancer risk following pernicious anaemia. Br J Cancer 1989; 59: 8103.
  • 20
    Chan JC, Liu HS, Kho BC, et al. Longitudinal study of Chinese patients with pernicious anaemia. Postgrad Med J 2008; 84: 64450.
  • 21
    Demmler K. Causes of death in patients treated for pernicious anemia. Med Klin 1966; 61: 5757.
  • 22
    Elsborg L, Andersen D, Bastrup-Madsen P. Gastrocamera screening in pernicious anaemia. With special reference to the occurrence of gastric polyps and cancer. Scand J Gastroenterol 1973; 8: 58.
  • 23
    Elsborg L, Andersen D, Myhere-Jensen O, et al. Gastric mucosal polyps in pernicious anaemia. Scand J Gastroenterol 1977; 12: 4952.
  • 24
    Hoffman NR. The relationship between pernicious anemia and cancer of the stomach. Geriatrics 1970; 25: 905.
  • 25
    Hsing AW, Hansson LE, McLaughlin JK, et al. Pernicious anemia and subsequent cancer. A population-based cohort study. Cancer 1993; 71: 74550.
  • 26
    Jorgensen J. The mortality among patients with pernicious anemia in Denmark and the incidence of gastric carcinoma among the same. Acta Med Scand 1951; 139: 47281.
  • 27
    Kokkola A, Sjöblom SM, Haapiainen R, et al. The risk of gastric carcinoma and carcinoid tumours in patients with pernicious anaemia. A prospective follow-up study. Scand J Gastroenterol 1998; 33: 8892.
  • 28
    Landgren AM, Landgren O, Gridley G, et al. Autoimmune disease and subsequent risk of developing alimentary tract cancers among 4.5 million US male veterans. Cancer 2011; 117: 116371.
  • 29
    Mellemkjaer L, Gridley G, Mller H, et al. Pernicious anaemia and cancer risk in Denmark. Br J Cancer 1996; 73: 9981000.
  • 30
    Mosbech J, Videbaek A. Mortality from and risk of gastric carcinoma among patients with pernicious anaemia. Br Med J 1950; 2: 3904.
  • 31
    Norcross JW, Monroe SE, Griffin BG. The development of gastric carcinoma in pernicious anemia. Ann Intern Med 1952; 37: 33843.
  • 32
    Schafer LW, Larson DE, Melton LJ 3rd, et al. Risk of development of gastric carcinoma in patients with pernicious anemia: a population-based study in Rochester. Minnesota. Mayo Clin Proc 1985; 60: 4448.
  • 33
    Siurala M, Eramaa E, Tapiovaara J. Pernicious anemia and gastric carcinoma. Acta Med Scand 1959; 164: 4316.
  • 34
    Sjöblom SM, Sipponen P, Järvinen H. Gastroscopic follow up of pernicious anaemia patients. Gut 1993; 34: 2832.
  • 35
    Svendsen JH, Dahl C, Svendsen LB, et al. Gastric cancer risk in achlorhydric patients. A long-term follow-up study. Scand J Gastroenterol 1986; 21: 1620.
  • 36
    Vannella L, Lahner E, Osborn J, et al. Risk factors for progression to gastric neoplastic lesions in patients with atrophic gastritis. Aliment Pharmacol Ther 2010; 31: 104250.
  • 37
    von Knorre G, Pechau KG. Late fate of patients with pernicious anemia. Z Gesamte Inn Med 1975; 30: 7016.
  • 38
    Wun Chan JC, Yu Liu HS, Sang Kho BC, et al. Pernicious anemia in Chinese: a study of 181 patients in a Hong Kong hospital. Medicine (Baltimore) 2006; 85: 12938.
  • 39
    Carmel R. Ethnic and racial factors in cobalamin metabolism and its disorders. Semin Hematol 1999; 36: 88100.
  • 40
    Ferlay J, Bray F, Pisani P, et al. GLOBOCAN 2002: Cancer Incidence, Mortality and Prevalence Worldwide. Lyon: IARC Cancer Base n.5 version 2 0, IARC Presss, 2004.
  • 41
    De Vries AC, Haringsma J, Kuipers EJ. The detection, surveillance and treatment of premalignant gastric lesions related to Helicobacter pylori infection. Helicobacter 2007; 12: 115.
  • 42
    Lahner E, Norman GL, Severi C, et al. Reassessment of intrinsic factor and parietal cell autoantibodies in atrophic gastritis with respect to cobalamin deficiency. Am J Gastroenterol 2009; 104: 20719.
  • 43
    Marignani M, Delle Fave G, Mecarocci S, et al. High prevalence of atrophic body gastritis in patients with unexplained microcytic and macrocytic anemia: a prospective screening study. Am J Gastroenterol 1999; 94: 76672.
  • 44
    Fong TL, Dooley CP, Dehesa M, et al. Helicobacter pylori infection in pernicious anemia: a prospective controlled study. Gastroenterology 1991; 100: 32832.
  • 45
    Haruma K, Komoto K, Kawaguchi H, et al. Pernicious anemia and Helicobacter pylori infection in Japan: evaluation in a country with a high prevalence of infection. Am J Gastroenterol 1995; 90: 110710.
  • 46
    Annibale B, Lahner E, Bordi C, et al. Role of Helicobacter pylori infection in pernicious anaemia. Dig Liver Dis 2000; 32: 75662.
  • 47
    Karnes WE Jr, Samloff IM, Siurala M, et al. Positive serum antibody and negative tissue staining for Helicobacter pylori in subjects with atrophic body gastritis. Gastroenterology 1991; 101: 16774.

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Authorship
  8. Acknowledgements
  9. References
  10. Supporting Information
FilenameFormatSizeDescription
apt12177-sup-0001-TableS1.docWord document72KTable S1. Electronic-search in Pubmed up to 15 September 2011.
apt12177-sup-0002-TableS2.docWord document79KTable S2. Quality assessment of included studies.
apt12177-sup-0003-TableS3.docWord document131KTable S3. Characteristics of the included studies.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.