Article Text

Download PDFPDF

Malignancies in cases with screening-identified evidence of coeliac disease: a long-term population-based cohort study
  1. S Lohi1,
  2. M Mäki1,2,
  3. J Montonen3,
  4. P Knekt3,
  5. E Pukkala4,
  6. A Reunanen3,
  7. K Kaukinen1,5
  1. 1
    Paediatric Research Center, Medical School, University of Tampere, Tampere, Finland
  2. 2
    Department of Paediatrics, Tampere University Hospital, Tampere, Finland
  3. 3
    National Public Health Institute, Helsinki, Finland
  4. 4
    Finnish Cancer Registry, Helsinki, Finland
  5. 5
    Department of Gastroenterology and Alimentary Tract Surgery, Tampere University Hospital, Tampere, Finland
  1. Dr K Kaukinen, Medical School, Building FM3, FIN-33014 University of Tampere, Finland; katri.kaukinen{at}


Background and aims: The association between diagnosed coeliac disease and malignancy has been established. The present study was conducted to determine whether previously unrecognised and thus untreated adults with screening-identified evidence of coeliac disease carry an increased risk of malignancies.

Methods: A Finnish population-based adult-representative cohort of 8000 individuals was drawn in 1978–1980. Stored sera of the participants with no history of coeliac disease or any malignancy were tested for immunoglobulin A (IgA) class tissue transglutaminase antibodies (Eu-tTG) in 2001. Positive sera were further analysed by another tissue transglutaminase antibody test (Celikey tTG) and for endomysial antibodies (EMAs). Malignant diseases were extracted from the nationwide database and antibody-positive cases were compared with negative cases during a follow-up of nearly 20 years.

Results: Altogether 565 of all the 6849 analysed serum samples drawn in 1978–80 were Eu-tTG positive. In further analyses, 202 (2.9%) of the participants were Celikey tTG positive and 73 (1.1%) were EMA positive. The overall risk of malignancy was not increased among antibody-positive cases in the follow-up of two decades; the age- and sex-adjusted relative risk was 0.91 (95% CI 0.60 to 1.37) for those who were Celikey tTG positive and 0.67 (95% CI 0.28 to 1.61) for those who were EMA positive.

Conclusions: The prognosis of adults with unrecognised coeliac disease with positive coeliac disease antibody status is good as regards the overall risk of malignancies. Thus, current diagnostic practice is sufficient and there is no need for earlier diagnosis of coeliac disease by mass screening on the basis of the findings of this study.

View Full Text

Statistics from

Coeliac disease, which is induced by ingestion of cereal gluten, is a chronic autoimmune-mediated disease with both intestinal and extraintestinal manifestations. Screening studies have revealed up to 1–2% seroprevalence of coeliac disease in both Europe and the USA.14 However, up to 75–90% of all cases remain unrecognised due to absent or atypical symptoms.16

According to follow-up studies, patients with diagnosed coeliac disease are at an increased risk of mortality and malignancies.7 The risk of malignancy over all sites has been shown to be two- to fivefold in studies published in the 1970s and 1980s.810 However, in recent studies the association between diagnosed coeliac disease and malignancy of any type has been much lower than previously reported, being at the most 1.3-fold.1116 There are solid data to suggest that long-term adherence to a gluten-free diet will reduce the incidence of complications such as malignancies.7 10 However, a majority of diseased individuals still remain unrecognised and untreated.5 6 17 18 We do not know whether these apparently clinically silent unrecognised cases also carry an increased risk of coeliac disease-related complications and, thus, whether the healthcare system should recognise and treat them with a gluten-free diet during the early stages of the disease.

To elucidate this issue, we carried out a cohort study with a follow-up time of nearly 20 years and used a Finnish national population-representative adult cohort of 8000 people gathered in 1978–1980. Coeliac autoantibody-positive cases in the cohort were regarded as most probably representing unrecognised coeliac disease in this study. All malignant diseases in the cohort were extracted from the Finnish cancer registry, and the occurrence of any malignancies was compared between antibody-positive and antibody-negative subjects throughout the follow-up period.


Study population

This population-based follow-up-study took advantage of the Mini-Finland Health Survey, which was carried out in 1978–1980. Details of the study design and the baseline results are extensively reported elsewhere.19 20 In brief, a nationally representative sample of 8000 persons has been drawn of the population aged ⩾30 by the stratified two-stage cluster sampling design planned by Statistics Finland. The second stage of sampling was performed in 40 areas in different parts of the country. The participants attended a health examination, which included interviews, questionnaires, drawing of blood samples and a clinical examination by a doctor. The participation rate was 90% (n = 7217), and 87% (n = 6993) of the cohort had sera available for the purposes of this study in 2001 (fig 1). Cases having a previous diagnosis of coeliac disease or dermatitis herpetiformis (n = 3), or any malignancy (n = 141) in 1978–1980, when the sera were drawn and the follow-up started, were excluded from the analysis, yielding 6849 (mean age 51, range 30–95, females 3680) participants for this study (fig 1).

Figure 1

Flow chart of the present study. Eu-tTG, immunoglobulin A (IgA) class tissue transglutaminase antibody (Eu-tTG umana IgA); Celikey tTG, IgA class tissue transglutaminase antibody (Celikey Tissue Transglutaminase IgA Antibody Assay); EMA, IgA class endomysial antibody. An asterisk indicates cases with any malignancy (n = 141) or previous coeliac disease or dermatitis herpetiformis diagnosis (n = 3)

All participants gave informed consent. The Ethical Committee of Tampere University Hospital, Tampere, Finland, approved the study protocol.

Measurement of antibodies

The previously collected blood samples were stored at −20°C for later analysis. In 2001, all 6849 serum samples were analysed for immunoglobulin A (IgA) class tissue transglutaminase antibodies (Eu-tTG umana IgA; Eurospital, Trieste, Italy (abbreviated as Eu-tTG in this sudy)). Positive sera were further analysed using both another IgA class tissue transglutaminase antibody kit (Celikey Tissue Transglutaminase IgA Antibody Assay; Pharmacia Diagnostics, Uppsala, Sweden (abbreviated as Celikey tTG)), and a test for IgA endomysial antibodies (EMAs). Both commercial tissue transglutaminase antibody kits use human recombinant tissue transglutaminase as antigen and the results are given in arbitrary units (AU). The cut-off point for the Eu-tTG was 7.0 AU/ml and for the Celikey tTG 5.0 AU/ml as instructed by the manufacturers. The EMAs were defined by a standardised and validated indirect immunofluorescence method, and a characteristic staining pattern at a serum dilution of 1:⩾5 was considered positive.21 22

Due to the unexpectedly high percentage of Eu-tTG positivity in the sera of the Mini-Finland survey collected 22 years earlier, we also randomly selected 128 (one in 50) Eu-tTG-negative serum samples and tested them for Celikey tTG and EMAs. None of the them was positive for Celikey tTG or EMAs (fig 1).6

Malignancies and possible confounders

Malignant diseases were identified by linking the personal identification codes with records from the nationwide database of the Finnish Cancer Registry, which includes >99% of incident cases diagnosed in Finland since 1953, and has been shown to be a valuable source of information as the data are of good quality with acceptably low false-positive and false-negative discrepancy rates.23 24 Follow-up commenced the day the blood samples were drawn, and the study subjects were followed-up for a maximum of 19 years until the occurrence of cancer, death or the end of 1996, whichever came first, yielding a total follow-up of 103 815 person years. The most common cancers in Finland and cancers previously shown to be associated with coeliac disease were analysed (lymphomas and breast, lung, prostate and gastrointestinal cancers).

Information on age, sex, smoking, body mass index, alcohol consumption, physical activity, consumption of bread, education, number of births and menopausal status was extracted for adjustment purposes.

Statistical analysis

A Cox regression model was applied to estimate relative risks (RRs) and 95% CIs for malignancy comparing antibody-positive with antibody-negative participants. Celikey tTG-positive cases were divided into tertiles to compare relative risks at different antibody levels. The possible confounding effects of age, sex, body mass index, smoking status, alcohol consumption, physical activity, consumption of bread and education were assessed using a series of multivariable models. Menopausal status and number of births were also adjusted for when analysing the association between coeliac disease and breast cancer. We fitted multiplicative interaction terms to assess possible interactions between antibody status and age, sex, smoking and body mass index. The analyses were performed on SAS 8.02 (SAS Institute, Cary, North Carolina, USA).


Altogether, 565 of the total 6849 analysed serum samples were Eu-tTG positive (fig 1). A total of 202 (2.9%, 129 females, mean age 59 years) of Eu-tTG positive cases were also Celikey tTG positive and, correspondingly, 73 (1.1%, 52 females, mean age 50 years) were EMA positive. The subjects positive for Celikey tTG were older and consumed more alcohol than those who were negative. Otherwise, no statistically significant differences across antibody status were detected (table 1).

Table 1 Association of personal characteristics with tissue transglutaminase antibodies (Celikey tTG) and endomysial antibodies (EMA)

During surveillance of the participants in this study, 694 (10.1%) developed a malignancy of some type. Those who developed cancer during the follow-up were older and were more likely to be men or smokers (data not shown).

Coeliac autoantibody positivity did not increase the overall risk of malignancy (table 2). The multivariate adjustment (age, sex, body mass index, smoking, physical activity, consumption of bread and alcohol consumption) did not change the risk level (0.95, 95% CI 0.62 to 1.44, p = 0.80) among Celikey tTG-positive cases. No association was found in different levels of tissue transglutaminase antibodies or after exclusion of EMA-positive individuals from the tissue transglutaminase-positive cases (data not shown).

Table 2 Age- and sex-adjusted relative risk of different cancers between persons with positive and negative tissue transglutaminase antibodies (Celikey tTG) and endomysial antibodies (EMA)

However, EMA positivity was statistically significantly associated with an increased risk of lymphoproliferative disease (table 2). When non-Hodgkin’s lymphomas were considered separately, the age- and sex-adjusted relative risk of this malignancy among EMA-positive cases was 6.43 (95% CI 1.52 to 27.22, p = 0.05, n = 2) and the corresponding risk for Celikey tTG-positive individuals was 2.92 (95% CI 0.87 to 9.74, p = 0.13, n = 3). The duration between known antibody positivity and the diagnosis of non-Hodgkin lymphoma varied between 6 and 14 years, and non-Hodgkin lymphoma had no specific site predilection (table 3). There were no enteropathy-associated T cell lymphomas. In addition, the age- and sex-adjusted relative risk of carcinoma of the oesophagus was increased among Celikey tTG-positive cases, being 7.48 (95% CI 2.06 to 27.25, p = 0.01, n = 3).

Table 3 Characteristics of non-Hodgkin’s lymphoma (NHL) and carcinoma of the oesophagus in coeliac autoantibody-positive cases

We also adjusted the relative risks for non-Hodgkin’s lymphoma and carcinoma of the oesophagus for alcohol consumption, as this differed between antibody-positive and -negative participants (table 1); the risks remained virtually the same after the adjustment. No statistically significant interactions between any of the potentially effect-modifying factors (age, sex, smoking and body mass index) and antibody status were noted in the prediction of malignancy of any type (data not shown).


Several studies have investigated the association between diagnosed and treated coeliac disease and malignancies.816 We still do not know whether apparently asymptomatic unrecognised coeliac disease cases are at an increased risk of cancers. We have carried out a cohort study of the association of coeliac autoantibodies and malignancies with a follow-up time of nearly 20 years. The majority of previously unrecognised antibody-positive cases would most probably have remained undiagnosed throughout the follow-up period, as, according to our previous data from Finland, 74% of coeliac disease cases still remained unrecognised in 2000.6 Our study design enabled us to approach a hitherto unexplored issue, the prognosis of the unrecognised part of the coeliac population with regard to any cancers.

The present study showed that there is no additional risk of overall malignancy among untreated adults with screening-identified evidence of coeliac disease in a follow-up of nearly 20 years. In the majority of recently published studies the association between diagnosed coeliac disease and malignancy of any type has likewise not been found.1116 In contrast, the risk of malignancy over all sites was increased in studies published in the 1970s and 1980s.810 The decrease in the risk of malignancy during the last few decades may be due to improved diagnostic activity, the increased number of coeliac disease cases with mild symptoms over time and, thus, early commitment to a gluten-free diet.10 As also discussed by Catassi and associates,25 a detection bias might also have caused overestimation of the risk of malignancy in the earliest studies. The likelihood of detecting an occult or overt malignancy in individuals with coeliac disease may be higher compared with the corresponding likelihood in a control group due to more careful examination of the diseased cases. There are no previous follow-up studies comparable with ours concerning the association of unrecognised coeliac disease with malignancies. However, the same issue has been approached from a different point of view with inconclusive results. Metzger and colleagues studied mortality among a tissue transglutaminase antibody-positive population and found a 3.6-fold excess of cancer as the cause of death during 8 years of follow-up.26 In contrast, no increased risk of non-Hodgkin’s lymphoma could be detected in a small subgroup of screen-detected silent coeliac disease cases in a case–control study design.27 We studied previously unrecognised coeliac autoantibody-positive cases and, as they most probably had a mild clinical picture, the risk of complications such as malignancies seemed to remain low.

However, earlier unrecognised coeliac autoantibody-positive cases may still carry an increased risk of specific malignancies such as non-Hodgkin’s lymphoma and carcinoma of the oesophagus. The number of detected malignancies remained low regardless of the follow-up time of nearly 20 years, and thus caution is warranted in interpretation of the results. However, our results are supported by earlier findings whereby diagnosed coeliac disease patients have repeatedly been at an increased risk of the same specific cancers.10 12 25 27

Since 2002 all coeliac disease or dermatitis herpetiformis patients in Finland have had a right to financial assistance from the Social Insurance Institution provided the diagnostic criteria for the diseases are fulfilled. According to the register, only four coeliac autoantibody-positive cases had received a correct diagnosis by the end of the follow-up period and only two of them followed a gluten-free diet (unpublished data). As only patients alive in 2002 were included in the register, there may be a few diagnosed cases we could not detect. Nonetheless, the number of diagnosed and treated cases during the follow-up seems to be low and, thus, the major part of the unrecognised coeliac autoantibody-positive cases in 1978–1980 had also remained unrecognised and untreated throughout the follow-up period.

According to the literature, IgA class tissue transglutaminase and EMAs are valid tests for coeliac disease. The pooled specificities of these antibodies have been reported to approach 100% and sensitivities to be mainly >90% in adult populations.28 29 As a result of imperfect sensitivity there may be some false-negative coeliac disease cases in our cohort due to either IgA deficiency or a total lack of coeliac autoantibodies, thus possibly slightly diluting the real difference. However, these autoantibodies have previously been used in large screening studies in the USA and Europe,1 3 4 and some centres even use serology as a diagnostic tool without intestinal biopsy.30 The detection of a surprisingly high frequency of Eu-tTG positivity possibly due to long-term storage has been discussed previously.6 For clarity and due to the low specificity of the test, we do not report the results based on that test though the results were parallel with the reported results using other antibody tests (data not shown). The problem with old sera could theoretically be fully obviated by a prospective study design, which would, however, take decades to yield information. Additionally, it might be unethical to follow cases with apparent coeliac disease for several years without treatment. Nonetheless, the validity of our conclusions was strengthened as we obtained parallel results independently of the antibody used.

In conclusion, the overall risk of malignancies was not increased among coeliac autoantibody-positive cases probably representing unrecognised coeliac disease. Thus early diagnosis of coeliac disease through serological mass screening would not be beneficial in improving the prognosis of these antibody-positive cases as regards malignancies.


View Abstract


  • Funding: The Coeliac Disease Study Group is supported by grants from the Competitive Research Funding of Pirkanmaa Hospital District, the Emil Aaltonen Foundation, the Foundation for Paediatric Research, the Yrjö Jahnsson Foundation, the Finnish Coeliac Society and the Academy of Finland, Research Council for Health. The present study was also funded by the Commission of the European Communities in the form of the Research and Technology Development programme “Quality of Life and Management of Living Resources” (QLRT-1999-00037), “Evaluation of the Prevalence of Coeliac Disease and its Genetic Components in the European Population”. The study does not necessarily reflect the current views or future policies of the Commission of European Communities. The authors’ work was independent of the funders.

  • Competing interests: None.

  • Ethics approval: The Ethical Committee of Tampere University Hospital, Tampere, Finland, approved the study protocol.

  • Patient consent: Obtained.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles