Incidence of gastric B-cell lymphomas: a population-based study in Germany

A. Ullrich1,2,+, W. Fischbach3 and M. Blettner1

1 School of Public Health, Department of Epidemiology and Medical Statistics, University of Bielefeld, Bielefeld; 2 World Health Organization Geneva Programme on Cancer Control, Geneva, Switzerland; 3 Medizinische Klinik II, Klinikum Aschaffenburg, Aschaffenburg, Germany

Received 2 July 2001; revised 18 December 2001; accepted 18 January 2002


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Background:

While the clinical and experimental knowledge concerning gastric lymphomas is increasing, there is a scarcity of epidemiological data.

Patients and methods:

A population-based sample of patients in Franconia and Saarland in Germany was collected from a clinical trial, hospital archives and a cancer registry.

Results:

Over a period of 3 years, 94 patients with primary gastric lymphoma were recorded out of a total population of 3.5 million. The standardised incidence rates in Saarland and Franconia were 0.7 and 0.8 cases per 100 000, respectively. Patients were predominantly from higher age groups (mean age 62.1 years) and the incidence in men was slightly more than in women (P <0.03). The distribution of histological subtypes in Franconia was as follows: marginal zone B-cell lymphomas (MZBL), 58%; diffuse large-cell B-cell lymphoma (DLBL), 33%; and mixed forms, 9%. Helicobacter pylori could be detected histologically in 84% of all cases, 95% of MZBL cases and 68% of DLBL cases.

Conclusions:

Incidence rates of gastric lymphoma in Germany were similar to that in other European countries, except England, where rates are lower. The subtype-specific differences of H. pylori infection rates could be due to differences in carcinogenesis or to secondary changes during malignant transformation.

Key words: diffuse large-cell B-cell lymphoma, epidemiology, gastric non-Hodgkin’s lymphoma, Helicobacter pylori infection rates, incidence rates, marginal zone B-cell lymphoma


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
There is increasing clinical and experimental knowledge but little epidemiological data concerning primary gastric lymphomas, the most frequent of which is gastrointestinal non-Hodgkin’s lymphoma (NHL). There are only five population-based studies on primary gastric lymphomas, extracted from regional registers of haematopoetic malignancies [15], and two studies on pooled data from different cancer registers [6, 7]. The results show a rather homogenous picture (incidence rates between 0.41 and 0.85 per 100 000; Table 1) with the exception of England, where the rates are below 0.3 per 100 000. However, comparisons are hampered by the fact that different selection procedures and methods of standardisation have been used in these studies.


View this table:
[in this window]
[in a new window]
 
Table 1. Population-based studies on gastric non-Hodgkin’s lymphomas
 
Despite convincing evidence that Helicobacter pylori is an important risk factor for gastric lymphoma [8], it is still un-clear whether H. pylori infection is mandatory for the carcinogenesis of all subtypes of gastric lymphoma [9]. Marginal zone B-cell lymphomas (MZBL) and diffuse large-cell B-cell lymphomas (DLBL) are the most frequent subtypes of gastric NHL following the terminology of the revised European–American (REAL) classification of lymphoid neoplasms of NHL [10] and the more recent WHO classification of NHL [11]. From clinical case series, a wide range of H. pylori infection rates (48–100%) have been reported [1214], without specifying the histological subtypes. To our knowledge, infection rates are not available from population-based samples.

We carried out a study to determine incidence rates and clinical features, such as histological type and H. pylori status, of gastric lymphoma in Germany from a population-based sample of affected patients. The study covered a total population of 3.5 million in the state of Saarland and the districts of Lower and Upper Franconia, the latter situated in the state of Bavaria (Figure 1).



View larger version (28K):
[in this window]
[in a new window]
 
Figure 1. Regional study areas and data sources. *, Department of Pathology, Community Hospital of Bayreuth, Upper Franconia; **, Department of Pathology and department of Internal Medicine, University Hospital Wuerzburg, Lower Franconia; §, Cancer Register of the State of Saarland, Saarbruecken. Study centre of the German–Austrian Multicentre Study on Gastrointestinal Lymphomas[15].

 

    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Data were extracted from different sources in order to have a population-based sample of patients with a high degree of completeness. In the Franconian districts, we collected histological records of patients affected by gastric lymphomas from the German–Austrian Multicentre Study on Gastric Lymphoma (multicentre study) [15] and from the clinical archives of the Departments of Pathology at the University of Wuerzburg and the Community Hospital of Bayreuth. The multicentre study recruited, from March 1993 to February 1996, a total of 266 newly diagnosed patients with primary gastric lymphoma at 163 co-operating clinical centres in Germany and Austria. Of the 266 patients, 28 were residents of Upper and Lower Franconia and were included in our study. The Franconian districts were chosen as study regions because the co-operating centres were clustered in this area. Secondly, the University of Wuerzburg and the Community Hospital of Bayreuth, which are also situated in this area, are two major referral centres for gastric lymphomas in Germany. Patients not eligible for the multicentre study, due to an age >75 years, primary nodal or HIV-associated lymphoma, or prior or concomitant malignancies, could be identified from the clinical archives of the two Departments of Pathology. Due to the recruitment method used in the multicentre study and the referral system of both Departments of Pathology, there was geographically overlapping information for Lower and Upper Franconia. Thus, our method of data collection provides a high rate of case ascertainment.

The quality standards of the histological diagnoses in the Franconian data sources were of the same high level. Patients in the multicentre study were only accepted for inclusion if the central histological review of the Department of Pathology at the University of Wuerzburg confirmed the collaborating centres’ initial diagnosis of primary gastric lymphoma. Both Departments of Pathology collaborated closely in the diagnosis of gastrointestinal lymphomas and both followed the REAL classification of NHL [10]. All cases were reviewed by H.-K. Müller-Hermelink (Department of Pathology, University of Wuerzburg) or by Stolte (Department of Pathology, Klinikum Bayreuth). Patients were included when MZBL or DLBL had been diagnosed and confirmed. Gastric lymphomas consisting of low- and high-grade components, as described and classified by de Jong (type B and type C) [16], were also included. Copies of the patients’ histological records were provided by the Departments of Pathology as well as the multicentre study archive. From these records clinical data were abstracted, such as H. pylori infection status by histological diagnosis. The data from the different Franconian sources were compared to eliminate duplication.

In order to obtain incidence rates for gastric lymphomas from another region, we analysed data provided by the Saarland Cancer Register, an official population-based cancer register which covers the state’s total population of 1.08 million. There are multiple, partly overlapping sources of notifications for this register, particularly from clinicians and pathologists. The rate of case ascertainment for this register has been estimated to be high [17]. Data on patients affected by gastric lymphomas have been available since January 1994; therefore, a 3-year time period from January 1994 to December 1996 was chosen for case recruitment. In order to determine the quota of gastric lymphoma within the whole group of NHL, we requested data from the Saarland Cancer Register for the same time period. All cases of NHL classified according to the International Classification of Diseases (ICD) 9 [18] coded 200 or 202 were included.

The demographic data of the catchment population by age group for Lower and Upper Franconia were published in the statistical yearbooks of Bavaria [19]. In the case of Saarland, these data were provided by the regional cancer register.

In order to compare the H. pylori infection rates of our patient study sample with the general German population, we referred to a representative sample from western Germany for which H. pylori testing had been performed [20].

Statistical analysis
All incidence rates were referred to as the number of cases per 100 000 persons at risk. Age-specific incidence rates were calculated separately for Lower and Upper Franconia and Saarland based on the mean population of the corresponding study periods. Owing to the small number of patients only four age groups were chosen. Based on pooled data from all study regions age-specific incidence rates were calculated in 10-year intervals. The corresponding 95% confidence intervals (CIs) were computed by means of the confidence limit factors for estimates of Poisson- distributed variables [21]. Age-standardised incidence rates were calculated by direct methods based on the New European standard population and the World standard population [22]. Gender differences of regional age-specific incidence rates were tested by stratified contingency tables, referring to the procedures proposed by Breslow and Day [23] using the StatXact-4 statistical package (Cytel Software Corporation, Cambridge, MA). As higher incidence rates for gastrointestinal lymphomas have been reported in men [24], a one-sided exact P value was considered.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
During a 3-year period, from March 1993 to February 1996, chosen for our study, 66 patients with primary gastric lymphoma were found in Lower and Upper Franconia. In Saarland, 28 patients were recorded in the regional Saarland Cancer Register from January 1994 to December 1996 out of a total of 357 cases of NHL (7.8%). The demographic and clinical features of the Franconian sample are presented in Tables 2 and 3. Patients were predominantly from the higher age groups, with a mean age of 62.1 years; the age range was considerably larger (30–89 years). The mean age of men (57.1 years) was lower than that of women (68.0 years). The male to female ratio was 1.14. Marginal zone B-cell lymphoma were diagnosed in 58% of patients, DLBL in 33% and mixed forms in 9%. Helicobacter pylori could be detected histologically in 84% of all cases (95% CI 75% to 93%), in 95% (95% CI 91% to 99%) of MZBL cases and in 68% (95% CI 41% to 95%) of DLBL cases.


View this table:
[in this window]
[in a new window]
 
Table 2.  Patient characteristics in Upper/Lower Franconia sample
 

View this table:
[in this window]
[in a new window]
 
Table 3.  Helicobacter pylori prevalence in Upper/Lower Franconia sample
 
The overall age-specific incidence rates calculated from pooled data from the study areas in Franconia and Saarland are presented in Figure 2. A clear increase with age can be seen. Below 40–49 years of age, age-specific rates do not exceed 1 in 100 000. In the age group 70–79 years of age, rates increase in men up to 3.94 in 100 000, although CIs are large. In women, age-specific incidence rates appear slightly lower, but CIs of male and female rates largely overlap. The shape of the curves are rather similar to the curve based on pooled data for the whole group of NHL in Saarland. The overall age-standardised incidence rate (European standard population [22]) for men was 0.86 and for women was 0.70.



View larger version (23K):
[in this window]
[in a new window]
 
Figure 2. (Top) Overall age-specific incidence rates of non-Hodgkin’s lymphomas in Saarland, Germany and (Bottom) age-specific incidence rates of primary gastric lymphomas in Franconia and Saarland, Germany.

 
The regional age-specific and standardised incidence rates of gastric lymphomas and of NHL in the Saarland are presented in Table 4. The pattern of rising incidence with age can be observed in both regions. There are small regional variations of rates in the corresponding age groups due to the small number of patients. The crude regional total incidence rates vary between 0.86 (95% CI 0.39–1.63) and 0.91 (95% CI 0.57–1.37). As the corresponding CIs largely overlap, no significant regional differences in the occurrence of gastric lymphoma can be seen. Incidence rates for men are slightly higher than for women in all age groups and regions (P <0.03 if all regions are combined).


View this table:
[in this window]
[in a new window]
 
Table 4.  Regional incidence rates of primary gastric lymphomas and non-Hodgkin’s lymphoma (Saarland)
 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
We report for the first time the incidence rates for gastric B-cell lymphomas in Germany and population-based data on the clinical features of these pathological entities within the gastrointestinal NHL, defined according to the REAL classification [10]. Prerequisites for reliable epidemiological estimates are the clear definition of the catchment population and a high degree of completeness of case ascertainment. The catchment population consists of the populations at risk residing in the districts of Lower and Upper Franconia and the state of Saarland. The completeness of case ascertainment in the Saarland Cancer Register is estimated to be high because the different sources of information (from pathologists, clinicians and death certificates) overlap widely [17].

In Lower and Upper Franconia, case ascertainment is also based on overlapping sources of information from pathologists (University of Wuerzburg, Bayreuth) and clinicians (co-operating clinical centres of the multicentre study). All patients residing in Franconia diagnosed during the study period could be captured by both sources. The possibility that patients residing in Franconia were diagnosed and treated outside the districts cannot be excluded. The similarity of the estimates for incidence rates in our study for the Franconian districts and for Saarland indicates that the ascertainment rates in the study regions should be at the same high level. This assumes that regional incidence rates of gastric lymphomas in this restricted geographical context are the same.

There are small differences in the time frames of data collection between Franconia and Saarland. In the Saarland Cancer Register, site-specific coding for NHL was introduced in January 1994; therefore, data collection for this region only started 9 months later than for the multicentre study. The effect of this time frame difference was considered to be negligible for our data. The published time trend analyses of gastric lymphoma in Europe show a variation ranging between no change [1, 2] up to an annual change of 6.3% [3]. Assuming the latter rate for our study regions, no substantial influence could be expected for the validity of our estimates. Time trend analysis in our study was not performed because the observation time was too short and the number of cases too small.

Compared with the incidence rates derived from the epidemiological studies on gastric NHL (Table 1) and adjusted for the different underlying standard populations, the German rates (European standard population, women 0.70, men 0.86; World standard population, women 0.38, men 0.51, total 0.45) are very similar to the results in Denmark (European standard population, women 0.59, men 0.85) [1] and USA (World standard population, total 0.41) [7], but lower than in France (World standard population, women 0.44, men 0.76) [2]. In England the annual incidence of gastric NHL is even lower than in Germany (World standard population, women 0.20, men 0.24) [3]. These findings could be partly explained by different detection procedures, but also by geographic variations of risk factors such as H. pylori prevalence. Both these facts could also explain the differences in time trends [6].

With regard to gender differences, our results confirm the findings that the incidence in men is slightly higher than in women [2, 24]. This difference could indicate that gender-specific constitutional or behavioural factors could play an aetiological role in the development of gastrointestinal lymphomas. The gender differences concerning the mean age at diagnosis could be due to the same factors or sex-specific differences in the utilisation of medical care.

Based on histological findings, the overall H. pylori infection rate in our patient sample was 84% (95% CI 75% to 93%). Ten of 66 patients (missing data for three patients) were H. pylori negative at time of diagnosis. The H. pylori infection rate in the general population of West Germany, based on serological testing of a national random sample of 1834 participants, has been reported at 58% (95% CI 54% to 62%) in the age groups >50 years of age [21]. However, the difference in infection rates may be greater than we have found: H. pylori-negative patients at the time of diagnosis may have been exposed to this risk factor prior to or in earlier stages of the disease, H. pylori having been eliminated at the later stages [9]. It is probable, therefore, that lifetime exposure of patients was underestimated by the histological findings [25].

The infection rate in MZBL, calculated separately for the two histological subtypes, was 95% (95% CI 75% to 99%), higher than in the DLBL group [68% (95% CI 41% to 95%)]. The CIs are large due to the small number of cases. Although they partially overlap, the difference in infection rates may not be a random occurrence. Two hypotheses can be deduced. (i) In DLBL the secondary elimination of H. pylori could be more important than in MZBL due to less favourable local conditions in the stomach during the malignant transformation to high-grade lymphoma. (ii) H. pylori could also have different roles in the pathogeneses of the two subtypes of gastric lymphoma, explaining, in part, the observed difference in infection rates. Further studies on this topic should include serological findings, as they reflect a more valid assessment of the history of H. pylori infection in patients. Because H. pylori infection in the general German population is common but gastric lymphoma is a very rare disease, further research should focus on risk factors other than H. pylori. Some evidence concerning pesticides and solvents [26], dietary compounds [27] or infectious agents such as Helicobacter heilmanii [28] has been presented in the literature. Additionally, host factors could also affect gastric lymphoma risk and should be further addressed.


    Acknowledgements
 
The authors thank the following people for help in providing data: Maria-Elisabeth Göbeler-Kolve (member of the German-Austrian Gastrointestinal Lymphoma Study Group); Michael Vieth, Manfred Stolte (Department of Pathology, Klinikum Bayreuth); Axel Greiner, Hans-Konrad Müller-Hermelink (Department of Pathology, University of Würzburg); Christa Stegmeier (Saarland Cancer Registry). The authors thank Hajo Zeeb (School of Public Health, University of Bielefeld) and Emanuele Zucca (Oncology Institute of Southern Switzerland, Department of Medical Oncology, Bellinzona) for comments on earlier versions of this paper.


    Footnotes
 
+ Correspondence to: Dr A. Ullrich, NMH/PCC WHO, 1211 Geneva 27, Switzerland. Tel: +41-22-791-1292; Fax: +41-22-791-4297; E-mail: ullricha@who.int Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
1. d’Amore F, Brinker H, Gronbaek K et al. Non-Hodgkin’s lymphoma of the gastrointestinal tract: a population-based analysis of incidence, geographic distribution, clinicopathologic presentation features, and prognosis. Danish Lymphoma Study Group. J Clin Oncol 1994; 12: 1673–1684.[Abstract]

2. Ducreux M, Boutron MC, Picard F et al. A 15-year series of gastrointestinal non-Hodgkin’s lymphomas: a population-based study. Br J Cancer 1998; 77: 511–514.[ISI][Medline]

3. Gurney KA, Cartwright RA, Gilman EA. Descriptive epidemiology of gastrointestinal non-Hodgkin’s lymphoma in a population-based registry. Br J Cancer 1999; 79: 1929–1934.[ISI][Medline]

4. Otter R, Bieger R, Kluin PM et al. Primary gastrointestinal non-Hodgkin’s lymphoma in a population-based registry. Br J Cancer 1989; 60: 745–750.[ISI][Medline]

5. Lybeer ML, De Neve W, Vrints LW et al. Primary gastric non-Hodgkin’s lymphoma stage IE and IIE. Eur J Cancer 1996; 32A: 2306–2311.

6. Severson RK, Davis S. Increasing incidence of primary gastric lymphoma. Cancer 1990; 66: 1283–1287.[ISI][Medline]

7. Newton R, Ferlay J, Beral V, Devesa SS. The epidemiology of non-Hodgkin’s lymphoma: comparison of nodal and extra-nodal sites. Int J Cancer 1997; 72: 923–930.[ISI][Medline]

8. Parsonnet J, Hansen S, Rodriguez L et al. Helicobacter pylori infection and gastric lymphoma. N Engl J Med 1994; 330: 1267–1271.[Abstract/Free Full Text]

9. Bouzourene H, Haeflinger T, Delacretaz F et al. The role of Helicobacter pylori in primary gastric MALT lymphoma. Histopathology 1999; 34: 118–123.[ISI][Medline]

10. Harris NL, Jaffe ES, Stein H et al. A revised European–American classification of lymphoid neoplasms: a proposal from the International Lymphoma Study Group. Blood 1994; 84: 1361–1392.[Free Full Text]

11. Harris NL, Jaffe ES, Diebold J et al. World Health Organization classification of neoplastic diseases of the hematopoietic and lymphoid tissue: report of the Clinical Advisory Committee Meeting, Airelie House, Virginia, November 1997. J Clin Oncol 1999; 17: 3835–3849.[Abstract/Free Full Text]

12. Xu WS, Ho FC, Ho J et al. Pathogenesis of gastric lymphoma: the enigma in Hong Kong. Ann Oncol 1997; 8 (Suppl 2): 41–44.[Abstract]

13. Karat D, O’Hanlon DM, Hayes N et al. Prospective study of Helicobacter pylori infection in primary gastric lymphoma. Br J Surg 1995; 82: 1369–1370.[ISI][Medline]

14. Fischbach W, Jung T, Goebeler-Kolve M-E et al. Comparative analysis of the Helicobacter pylori status in patients with gastric MALT-type lymphoma and their respective spouses. Z Gastroenterol 2000; 38: 627–630.[ISI][Medline]

15. Fischbach W, Dragosics B, Kolve M-E et al. Primary gastric B-cell lymphoma: results of the German–Austrian Prospective Multicenter Study. Gastroenterology 2000; 119: 1191–1202.[ISI][Medline]

16. de Jong D, Boot H, Taal B. Histological grading with clinical relevance in gastric mucosa-associated lymphoid tissue (MALT) lymphoma: recent results. Cancer Res 2000; 156: 27–32.

17. Brenner H, Stegmaier C, Ziegler H. Estimating completeness of cancer registration in Saarland/Germany with capture–recapture methods. Eur J Cancer 1994; 11: 1659–1663.

18. World Health Organization. International Classification of Diseases, 9th edition. Geneva, Switzerland: WHO 1977.

19. Bayerisches landesamt für statistik und datenverarbeitung. Altersstruktur der bevölkerung bayerns. In Statistische Berichte des Bayerischen Landesamtes für Statisitik und Datenverarbeitung. Munich, Germany: 1994; 1995; 1996.

20. . Kliebsch U, Berg G, Lappus N et al. Prävalenz der infektion mit Helicobacter pylori in einer repräsentativen deutschen Bevölkerungsstichprobe der alten Bundesländer. In 43. Jahrestagung der GMDS. Bremen, Germany: MMV Medien & Medizin 1998; 471–474.

21. dos Santos Silva S. Cancer Epidemiology. Lyon, France: IARC 1999.

22. Parkin DM, Muir CS, Whelan SL et al. Cancer Incidence in Five Continents, Vol. VI. Lyon, France: IARC 1992.

23. Breslow NE, Day NE. Statistical Methods in Cancer Research. Vol. I: The Analysis of Case–Control Studies. Lyon, France: IARC 1980.

24. Cantor KP, Fraumeni, JF Jr. Distribution of non-Hodgkin’s lymphoma in the United States between 1950 and 1975. Cancer Res 1980; 40: 2645–2652.[Abstract]

25. Eck M, Schmauser B, Greiner A et al. MALT-type lymphoma of the stomach is associated with Helicobacter pylori strains expressing the CagA protein. Gastroenterology 1997; 112: 1482–1486.[ISI][Medline]

26. Hardell L, Axelson O. Environmental and occupational aspects on the etiology of non-Hodgkin’s lymphoma. Oncol Res 1998; 10: 1–5.[ISI][Medline]

27. Tavani A, Pregnolato A, Negri E et al. Diet and risk of lymphoid neoplasms and soft tissue sarcomas. Nutr Cancer 1997; 27: 256–260.[ISI][Medline]

28. Morgner A, Lehn N, Andersen LP et al. Helicobacter heilmannii-associated primary gastric low-grade MALT lymphoma: complete remission after curing the infection. Gastroenterology 2000; 118: 821–828.[ISI][Medline]