Skip to main content
  • Research article
  • Open access
  • Published:

Validation study of villous atrophy and small intestinal inflammation in Swedish biopsy registers

Abstract

Background

Small intestinal biopsy with villous atrophy (VA) is the gold standard for the diagnosis of celiac disease (CD). We validated VA (Marsh 3) and small intestinal inflammation without VA (Marsh 1+2) in Swedish regional biopsy registers.

Methods

All pathology departments in Sweden (n = 28) were searched to identify individuals with VA or duodenal/jejunal inflammation. The validation consisted of blinded examination of biopsy samples, manual review of biopsy reports, web surveys, and patient chart reviews of 121 individuals with VA and 39 with inflammation.

Results

We identified 29,148 individuals with VA and 13,446 individuals with inflammation. In a blinded examination, Swedish pathologists correctly classified 90% of biopsies with VA. Manual screening of 1,534 biopsy reports (performed by co-author JFL and a research assistant) found that comorbidity other than CD was rare. IBD was the most common comorbidity and occurred in 0.3% of biopsies with VA (1.6% in inflammation). Among 114 patients with VA and available data, 108 (95%) had a clinical diagnosis of CD. 79% of the validated individuals with VA and 64% of those with inflammation had documented gastrointestinal symptoms prior to biopsy. 88% of the validated individuals with VA had positive CD serology before their first biopsy. 172/180 (96%) of Swedish gastroenterologists and 68/68 (100%) of pediatricians perform a small intestinal biopsy in at least 9 out of 10 individuals prior to diagnosis of CD.

Conclusion

Regional biopsy data are feasible to identify individuals with CD and small-intestinal inflammation. The specificity of CD is high in villous atrophy.

Peer Review reports

Background

Small intestinal biopsy is the gold standard of celiac disease (CD) diagnosis [1]. In a patient with CD, the biopsy typically shows villous atrophy (VA; Figure 1). Treatment with a gluten-free diet should then lead to improvement in symptoms and intestinal morphology on control biopsy [2, 3].

Figure 1
figure 1

Photographs of the small intestine: classifcation of histopathology. a. Normal mucosa. b. Inflammation (intraepithelial lymphocytosis). c. Partial VA (CD). d. Total VA (CD). (Acknowledgement. Photomicrographs obtained from Prof. Åke Öst, earlier chairman of the Swedish National Steering Group for Small Intestinal Pathology).

We identified 29,148 individuals with VA and 13,446 individuals with small intestinal inflammation (called "inflammation" in this paper) through a search of all pathology departments of Sweden (n = 28).

The overall aim of the current study was to validate VA and inflammation in small intestinal biopsies obtained from Swedish biopsy registers and to evaluate if biopsy registers can be used as a data source for CD and inflammation. We specifically examined a) the accuracy of histopathological classification of small intestinal biopsies by Swedish pathologists; b) the proportion of biopsy reports with VA or inflammation where other disease than CD is mentioned; c) to what extent VA corresponds to a clinical diagnosis of CD; d) the characteristics (symptoms and CD serology) of 160 validated individuals with VA or inflammation; and e) the use of small-intestinal biopsy among Swedish gastroenterologists and pediatricians to diagnose CD.

Methods

Classification of biopsy data

Pathologists in Sweden, and elsewhere, will routinely examine the presence (or absence) of VA, crypts, the different layers of the intestine, and the number of intraepithelial lymphocytes (IELs) (Table 1) in small intestinal biopsies. Finally, the pathologist examines the brush border for evidence of infection. In Sweden, all small intestinal biopsies are classified according to the SnoMed system for diagnostic and register purposes. SnoMed ("Systematized Nomenclature of Medicine-Clinical Terms") is a multilingual clinical healthcare terminology http://www.ihtsdo.org/our-standards/. In Sweden, SnoMed codes are often complemented with a "KVAST/Alexander" code" that is communicated to the physician who performed the biopsy. From an international perspective, the most common classification system of CD is the Marsh system ("Marsh-Oberhuber") [4], although recently Corazza et al suggested an alternative classification system with improved degree of interobserver reproducibility [5]. Table 1 contains an overview of different histopathology classifications and lists the SnoMed codes used in this project to identify individuals with VA and inflammation. In this project, inflammation is equivalent to Marsh 1+2 and VA equivalent to Marsh 3. Swedish SnoMed classification does not allow distinction between Marsh 1 and Marsh 2.

Table 1 Small intestinal histopathology classifications – a comparison

Collection of biopsy data

Biopsy data were collected through computerized searches of all regional pathology departments in Sweden (28 out of 28 departments). These departments also included the private pathology department Medilab, which is responsible for the examination of biopsy samples from many private pathologists, as well as samples from the Gotland County. Medilab has also archived biopsy results from former private pathologists who have retired. Regional biopsy data in Sweden are primarily archived in two different computer systems "Safir" and "Sympathy". Local IT technicians were asked to perform searches for appropriate morphology codes and then save data on a) arrival date of the biopsy, b) Personal Identity Number, c) morphology and d) topography (T64 = duodenum; T65 = jejunum). Computerized searches were then exported to Microsoft Excel format and delivered to the researchers. In total we received 189 Microsoft Excel files from the 28 pathology departments. Each file was examined by JFL and LB. On one occasion, the computerized search was inadequate and had to be re-performed at the regional pathology department. JFL and LB then collaborated on cleaning the data and identifying data irregularities. Biopsy data were finally merged into one file.

Initially we collected all small intestinal biopsy samples (normal mucosa, inflammation and VA) and at the end of the data collection, identified individuals with VA and inflammation according to relevant SnoMed codes. Data collection took place between the 27th of October 2006 and the 12th of February 2008.

Validation of VA and inflammation

The accuracy of histopathological classification of small intestinal biopsies was surveyed in 1999, when the Swedish National Steering Group for Small Intestinal Pathology sent out small intestinal biopsy samples to 22 Swedish pathology departments. Each biopsy sample had first been categorized through consensus of the steering group (gold standard) and was then graded by blinded pathologists from the 22 pathology departments. This validation (distribution of biopsy samples and collection of results) was coordinated by the non-profit company Equalis http://www.equalis.se/. Data on the accuracy were obtained from Equalis and compiled by JFL.

We evaluated the specificity of VA and inflammation. We did so by estimating the proportion of biopsy reports (with VA or inflammation) where other disease than CD was mentioned. In total we had data on 56,176 unique biopsies (VA: 38,655; inflammation: 17,521) in 29,148 individuals with CD and 13,446 individuals with inflammation. A computerized search of the plain text of these biopsy reports yielded 1,534 biopsy reports where there were indications of comorbidity other than CD. Each of these biopsy reports (n = 1,534) was then manually screened by two independent reviewers (JFL and research assistant) to confirm or reject the presence of comorbidity. Discrepancies were resolved through a third review of the biopsy report in question.

The manual review of biopsy reports was then complemented with a review of patient charts from 160 individuals undergoing small intestinal biopsy (121 with VA and 39 with inflammation). These individuals were randomly selected from five counties contributing biopsy data (these counties in total supplied 5,317 individuals with VA and 2,215 with inflammation). These 160 validated individuals had been referred for biopsy from 45 different clinics or health care centers. We specifically looked at the clinical characteristics of patients with VA and inflammation and dietary treatment and dietary compliance in individuals with VA. We were able to characterize 118 individuals with VA and 39 individuals with inflammation with regards to clinical symptoms and laboratory measures. There were patient chart data on dietary compliance in 71–101 individuals with VA. Finally we obtained all available CD serology data on the 160 validated individuals described above in order to correlate CD serology with VA.

We also examined the clinical management of CD in Sweden. This included the use of small intestinal biopsy prior to diagnosis of CD, the histopathological examination of small intestinal biopsies by pathologists and the treatment of CD. For this purpose we carried out three web surveys. Potential survey respondents were identified through member registers, email lists and through the Swedish Association of Pathology.

To identify (adult) gastroenterologists, we matched the member register of the Swedish Gastroenterology association with the administrative register of Swedish physicians http://www.hsar.se/ and identified 282 active gastroenterologists with a postal address in Sweden and an e-mail address. These individuals were then sent information by post and email about a survey on CD in adults, as well as the web address of an 11-item-survey. The results of this survey have previously been published in a Swedish-language journal (Gastrokuriren).

We identified pediatricians with an interest in gastroenterology through two existing e-mail lists for Swedish pediatricians (one for CD and one for general pediatric gastroenterology, including IBD). 70 pediatricians were then emailed information about a survey on CD in children, as well as the web address of a pediatric 11-item-survey. The results of this survey have previously been published in a Swedish-language journal (Barnläkaren).

After permission from the Swedish-language journals Gastrokuriren and Barnläkaren, we included data on CD investigation and management in this paper[6, 7], since these Swedish journals are not indexed by Medline (or any other medical database) and contain no English-language abstract. Written permissions from Gastrokuriren and Barnläkaren have been separately uploaded to BMC Gastroenterology.

To define the histopathological examination of small intestinal biopsies, we emailed 25 pathology departments and asked them to answer an 18-item-survey. Although, 28 pathology departments contributed data to our study, 3 smaller pathology departments participating in our study formally had the same head pathologist as the university hospital of their health care region and referred us to the larger department for our survey.

Statistics

We calculated 95% confidence intervals (CI).

Ethics

The current study was performed as part of a larger project on complications in CD and inflammation. That study was approved by the Regional Ethical Review Board in Stockholm on the 4th of June 2006 (2006/633-31/4) with additional amendments (2007/747-32 and 2008/257-32).

For the validation of the 160 patients, we obtained specific permission from the National Board of Health and Welfare.

Results

We identified 381,043 small intestinal biopsies. Of these, 28,654 were duplicates (biopsies could be identified more than once if histopathology SnoMed codes indicated both inflammation and villous atrophy). Of the remaining 352,389 biopsies we excluded 986 due to data inconsistencies (including incorrect Personal Identity Number). The final 351,403 small intestinal biopsies, from which we identified individuals with VA and inflammation, represented 287,586 unique individuals. 240,922 individuals had only one small intestinal biopsy. Less than 1% had been biopsied on more than three occasions.

We identified 29,148 unique individuals with VA at some stage and 13,446 unique individuals with inflammation at some stage but never VA. The median year of biopsy with VA or inflammation was 1998, ranging from 1969–2008 in individuals with VA to 1970–2008 in individuals with inflammation. Except for the above description of the biopsy data collection, all data in this paper refer only to individuals with VA or inflammation.

Age and sex of individuals with a small intestinal biopsy

The majority of study participants were females; and adults predominated at first biopsy (Table 2). In individuals with VA 36.8% were children (≤ 15 years) at first positive biopsy, compared with 6.1% of those with inflammation.

Table 2 Characteristics of unique individuals at first positive small intestinal biopsy

Validation of VA and inflammation

Blinded classification of small intestinal biopsies

Ninety percent (90%; 95% CI = 87–94%) of biopsies with VA were correctly classified as defined by the National Steering Group of Small Intestinal Pathology (Table 3). The proportions of missing data were independent of the grading of the biopsy. Fifty-six percent (41–72%) of biopsies with inflammation were correctly classified.

Table 3 Accuracy of histopathological classification

Manual examination of biopsy reports

Our examination of 1,534 biopsy reports, found that small intestinal pathology other than CD, as mentioned in the biopsy report, was uncommon in VA or inflammation (Table 4). The most common comorbidity was IBD, occurring in 0.3% (0.2–0.3%) of biopsies with VA and 1.6% (1.4–1.7%) of biopsies with inflammation (see Additional File 1.

Table 4 Comorbidity in biopsy samples – results of manual examinations of 1,534 biopsy reports

CD among individuals with VA

According to patient chart data, VA had a high specificity for CD. 108 (95%; 91–99%) out of 114 patients with VA and available data had a clinical diagnosis of CD (Figure 2) [8]. In seven individuals (121 minus 114), the CD diagnosis could neither be confirmed, nor rejected (Figure 2). This was due to emigration, inconsistent SnoMed coding etc. In individuals where the diagnosis was rejected (n = 6), four biopsies (patients 1–4) had been misclassified in the relevant department of pathology. Patient 5 had VA, but received gluten-containing food after the diagnostic biopsy. A second upper endoscopy 18 years later found that patient 5 had a normal mucosa and the CD diagnosis was re-evaluated. In patient 6, a biopsy initially classified as "VA" was later re-examined (on request from the physician) and then found not to fulfill the criteria for VA (or CD).

Figure 2
figure 2

Overview of patient chart review. 95% of all individuals with VA had a clinical diagnosis of CD according to the patient charts (95% CI = 91–99%). This is consistent with data from an earlier Swedish study showing that out of 524 children with a positive first biopsy, 509 improved in their mucosa when starting on a gluten-free diet (corresponds to 97.1%)[8]. *Other sources: In two patients, we phoned the responsible physician and the diagnosis of CD could then be confirmed. In a third patient, no patient chart could be obtained. However, through the local chemistry laboratory, we obtained data on CD serology that showed positive endomysium and antigliadin antibodies prior to biopsy, and negative endomysium and antigliadin antibodies after biopsy. Also this patient was regarded as having CD.

Clinical characteristics of individuals with VA and inflammation

Clinical characteristics of the reviewed individuals are presented in Table 5. Most individuals presented with classical features, with diarrhea being the most common symptom in individuals with VA. Although, 64% (49–79%) of individuals with inflammation had some gastrointestinal symptom prior to biopsy, the single most common symptom/sign was anemia (31%; 16–45%). Folic acid deficiency was reported in 22% (15–30%) of individuals with VA, but only in 5% (1–17%) of individuals with inflammation.

Table 5 Clinical characteristics of individuals with VA and inflammation – patient chart review.

93% of individuals with VA (94/101; 88–98%) had received dietary information according to patient chart data. There were indications of low dietary compliance (not strict gluten-free diet) in 15/86 individuals with VA (17%; 9–25%). Five out of 71 patients (7%; 1–13%) did not respond satisfactorily to a gluten-free diet. Two of the patients with little response to a gluten-free diet however admitted low dietary compliance.

Two individuals (5%; 1–17%) with inflammation (and who had never had VA) had at some stage received information about gluten-free diet. None of these individuals had however received a clinical diagnosis of CD according to the patient chart.

In 71 (88%; 80–95%) out of 81 individuals with VA and available CD serology data, there was at least one positive antibody test (gliadin, endomysium or transglutaminase) prior to first biopsy (Table 6). After biopsy, most serology samples in the validated individuals with VA were negative (Table 6). Among 39 individuals with inflammation (and no VA), two (5%; 1–17%) had a positive CD serology prior to biopsy (none of them received a clinical diagnosis of CD according to patient charts).

Table 6 CD serology before and after first biopsy in individuals with VA

Clinical management of CD

Our web surveys found that most gastroenterologists and pediatricians biopsy their patients prior to CD diagnosis and that almost all patients with CD receive information about gluten free diet.

Among 282 available gastroenterologists, 184 (65%) gastroenterologists participated in our survey. 96% (91–98%) of Swedish gastroenterologists perform a small intestinal biopsy in at least 9 out of 10 patients prior to diagnosis of CD (Table 7). 93% (89–97%) inform at least 9 out of 10 patients with CD about gluten-free diet.

68 pediatricians (97% of those contacted) participated in our survey (Table 8). 100% (95–100%) of the pediatricians perform a small intestinal biopsy in at least 9 out of 10 patients prior to diagnosis of CD. 97% (90–100%) inform at least 9 out of 10 patients with CD about gluten-free diet.

Table 7 Management of CD among gastroenterologists
Table 8 Management of CD among pediatricians

Survey of pathology departments: evaluation of CD and inflammation

Among 25 contacted departments of pathology, 23 (92%) responded to our histopathology survey. 20/23 (87%; 73–100%) pathology departments use a Swedish histopathological classification system called "KVAST/Alexander", with 3 departments using the Marsh classification when they communicate pathology findings to the physician (Table 1). All pathology departments however also classify biopsy samples according to the international SnoMed classification system http://www.snomed.org/. 20/23 (87%; 73–100%) pathology departments reported that most pediatricians and gastroenterologists would send 2–3 tissue samples from the duodenum/jejunum (originating from one endoscopy session) for referral (1 (4%) reported that most gastroenterologist would send only 1 biopsy samples for microscopic examination; while 2 departments (9%) reported 4 or more samples to be routine). 19/23 (83%; 67–98%) pathology departments said that their histopathology report would be based on the most severe finding at microscopic examination, while 4 (17%) reported that the most common finding would form the basis of their pathology report. When asked what (minimum) villous-crypt ratio would be accepted as normal, 3:1 was used in most departments (18/23; 78%; 64–95%) (all other departments reported a 4:1 (13%) or a 2:1 (9%) ratio to be accepted as normal). None of the pathology departments requested a 5:1 ratio. 22/22 (100%; 85–100%) pathology departments used immunostaining for CD3 to detect IELs. 19/23 (83%; 67–98%) reported using 30 IELs per 100 as their cut-off for "increased IEL count", while 1 department reported using 25 and 3 departments used 20 as their cut-off. Other results of this survey are presented in Table 9.

Table 9 Management of CD in pathology departments (%)

Discussion

This paper describes the validation of VA and inflammation in small intestinal biopsies from the 28 pathology departments in Sweden. In Sweden, pathologists have traditionally divided CD-like lesions into a) VA and b) inflammation without VA, with further subdivisions into partial, subtotal and total VA. Inflammation (Table 1) corresponds to both Marsh 1 and Marsh 2 and grade A in the newly introduced classification by Corazza et al [5]. This means that Swedish biopsy registers do not distinguish between Marsh 1 and Marsh 2.

We identified 29,148 individuals with VA and 13,446 individuals with inflammation through computerized searches of all pathology departments in Sweden. In our validation, VA had a high specificity for CD, which is not surprising since other causes of VA than CD are rare in Western countries [9]. In our manual review of more than 1,500 biopsy reports, comorbidity such as IBD, giardiasis, helicobacter pylori-infection, gastric metaplasia or cancer was rare in VA or small intestinal inflammation (Table 4) (also for biopsy samples with inflammation, was other co-morbidity than CD uncommon. In "inflammation", IBD accounted for 1.6% of biopsies with small-intestinal inflammation). Our patient chart review found that 95% of biopsy reports with VA were consistent with clinical CD. Although, VA may be patchy and biopsy samples without atrophy may be misinterpreted as atrophic by the pathologist if the samples are wrongly oriented, earlier validation has shown that the vast majority (90%) of intestinal biopsies with VA are correctly classified by Swedish pathologists (Table 3). Although, not all pathologists responded to this survey (Table 3), missing responses were equally distributed between biopsies with normal mucosa, inflammation and VA. This suggests that missing responses were not biased towards biopsy samples that are difficult to evaluate since we then would have expected a higher rate of missing values in samples with inflammation. Pathologists also reported that almost 9 out of 10 pediatricians/gastroenterologists submit at least two biopsy samples from the small intestine (instead of just one sample) when CD is investigated. Multiple samples increase the likelihood of identifying VA. We conclude that biopsy data with VA may be used as a data source for CD, although comorbidity may be underestimated when evaluated from biopsy reports since not all comorbidity is mentioned in biopsy reports.

Unfortunately we did not have access to the referral notes from the physician. Such notes often contain additional clinical information about the reason for biopsy. With access to referral note data it is possible that we could have distinguished individuals with inflammation biopsied due to suspected CD from those biopsied due to other cause such as IBD. To some extent this can be done through manual screening of the running biopsy text, but it is still likely that inflammation in Swedish biopsy registers is a heterogeneous concept representing both CD and other diseases.

The sensitivity of diagnosed CD should be close to 100% through biopsy registers, since a duodenal/jejunal biopsy with VA is gold standard for CD diagnosis in Sweden [10, 11] (see also Tables 7 and 8). However, although data on CD management were obtained from almost all pediatricians surveyed, our response rate was lower among gastroenterologists investigating and treating adults (65%). This response rate is similar to that in a review by Asch et al (usually 5–6 out of 10 physicians respond to mailed surveys) [12]. Nevertheless, there is a risk of selection bias and we cannot exclude that our respondents were more interested in CD than the average Swedish gastroenterologist and that this influenced the results of our survey.

Until now, most large Swedish cohort studies of CD have used the Swedish National Hospital Discharge Register to identify patents with CD (e.g. [13–16]). Regional biopsy registers do however offer several advantages over the Hospital Discharge Register when it comes to identifying CD (and inflammation). Foremost among these advantages may be higher specificity and sensitivity for both CD and inflammation.

While the Swedish Hospital Discharge Register and other diagnostic registers, cannot be used to identify inflammation, biopsy registers constitute a means to identify tissue-verified inflammation. Data from the Swedish National Steering Group for Small Intestinal Pathology show that 56% of biopsy samples with inflammation are correctly classified. This is lower than for VA but may still be higher than when using the traditional Marsh classification system. Our data are in fact similar to those reported by Italian researchers using a simplified Marsh classification system [5]. The Italian researchers had earlier noted that the interobserver reproducibility was lower for Marsh 1 and Marsh 2 (especially for Marsh 2) [5] than for Marsh 3 (especially Marsh 3c). However, partial Kappa scores increased significantly when they merged Marsh 1 and Marsh 2 lesions into what they call "grade A". This concept (Grade A) is actually identical to "small intestinal inflammation" in Swedish biopsy registers. The traditional use of grading small intestinal inflammation into VA vs. inflammation (without VA) stems from the long-term use of VA as the basis of CD diagnosis in Sweden [2] Increasingly it has however been recognized that CD may be present in the absence of VA and that additional measures including genetic and immunological markers may be important in the investigation of CD ([17]. The current study had no data on either genetic or immunological markers. Previous research suggests that up to 40% of individuals with inflammation (as defined by increased IEL count) may suffer from gluten sensitive enteropathy [18], but as noted earlier our histopathology approach does not distinguish between CD-inflammation and non-CD-inflammation.

Another weakness of our methodological approach is that we have no data on the exact numbers and distribution of IELs in biopsies with inflammation. It has been suggested that a uniform distribution of IELs is increases the proportion of CD in patients with inflammation without VA [19]. The lack of exact number and distribution of IELs in biopsies with inflammation is a limitation of our study.

"Positive biopsy" for VA has been the recommended gold standard for CD for the last 30–40 years[2, 20]. In contrast, CD serology has only been generally available in Sweden since the mid-1990s. This means that using biopsy registers to identify individuals with CD allows for substantially longer follow-up of patients than if using CD serology data to identify patients with CD. Even though the sensitivity of e.g. tissue transglutaminase autoantibodies is high in total VA, it decreases to around 70% in partial VA (also CD) and is even lower in inflammation [21]. CD serology has a high specificity [22] for CD and a high negative predictive value. CD serology is therefore well suited to rule out the presence of CD. In contrast, the positive predictive value of CD serology has been disappointing, which is not surprising given that 99% of all Westerners do not have CD [23, 24] (only a minute decrease in specificity will then result in false-positive cases). In two recent papers, the positive predictive value of tissue transglutaminase antibodies varied between 29–76%[25, 26]. For the above reasons we chose to identify CD and inflammation through biopsy reports instead of serological data.

Symptoms and signs vary in CD [27–29]. "Classical CD" presents with chronic diarrhea, abdominal pain, other gastrointestinal symptoms and weight loss/growth retardation [30] and is still common in children [27]. Increasingly however, CD is detected in children with non-classic symptoms, or through screening of first-degree relatives and high-risk individuals (e.g. those with type 1 diabetes) [31]). In adults, diarrhea is probably the dominating symptom (i.e. the classic form), although silent CD [30] is also common [32]. One of the limitations of biopsy registers, is that they contain no data on symptoms. We therefore characterized 121 patients with VA and 39 patients with inflammation through patient chart reviews and collection of CD serology data. Most of our patients with VA presented with classical symptoms (diarrhea was the most common symptoms). One in three patients with VA and inflammation suffered from anemia at time of biopsy. Our data on anemia and diarrhea in VA are thereby consistent with earlier findings in CD[27, 30, 33]. Heredity for CD was found in 12% of individuals with VA and in 5% of individuals with inflammation. Although, we do not know of any study examining the heredity for CD in individuals with inflammation without VA, Dube et al [24] recently compiled data on CD heredity, with a pooled prevalence of biopsy-verified CD of 16% among first-degree relatives to CD patients. It is also noteworthy that 88% of the 81 validated patients with available CD serology data and VA in our study had at least one positive CD serological marker prior to first biopsy. Although, most patients with VA had negative CD serology after first biopsy, not all did. Positive serology after biopsy may be explained by a number of factors including low compliance, serology performed only shortly after biopsy, or being performed at time of gluten provocation. In addition, a number of patients with VA may have persistently abnormal CD serology despite gluten-free diet.

Among patients with VA and available data, we also characterized dietary compliance. This review found indications of low dietary compliance in 17% of individuals with VA. This is consistent with earlier data [34–37], although it should be remembered that our data on dietary compliance were retrospectively collected through patient charts. Another weakness of our data on dietary compliance is that the follow-up differed between patients, as did the time between CD diagnosis and evaluation of dietary compliance. Finally, there is a risk that the basis of compliance assessment differed between physicians and dieticians; and that patients without information on dietary compliance differ from those with such information. Only 2/39 validated patients with inflammation in our study had received information about gluten-free diet and none of them had received a clinical diagnosis of CD by their physician. This is not surprising, given that Swedish patients with small-intestinal inflammation by tradition do not receive a gluten-free diet.

Conclusion

In conclusion, we merged data from all 28 pathology departments in Sweden and identified 29,148 individuals with VA and 13,446 individuals with small-intestinal inflammation. Our validation shows that it is feasible to identify individuals with CD and small-intestinal inflammation using regional biopsy data. The specificity of CD is high in villous atrophy.

Abbreviations

CD:

celiac disease

IEL:

intraepithelial lymphocytes

VA:

villous atrophy.

References

  1. AGA Institute Medical Position Statement on the Diagnosis and Management of Celiac Disease. Gastroenterology. 2006, 131: 1977-1980. 10.1053/j.gastro.2006.10.003.

  2. Revised criteria for diagnosis of coeliac disease. Report of Working Group of European Society of Paediatric Gastroenterology and Nutrition. Arch Dis Child. 1990, 65: 909-911. 10.1136/adc.65.8.909.

  3. Kupper C: Dietary guidelines and implementation for celiac disease. Gastroenterology. 2005, 128: S121-127. 10.1053/j.gastro.2005.02.024.

    Article  CAS  PubMed  Google Scholar 

  4. Marsh MN: Grains of truth: evolutionary changes in small intestinal mucosa in response to environmental antigen challenge. Gut. 1990, 31: 111-114. 10.1136/gut.31.1.111.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  5. Corazza GR, Villanacci V, Zambelli C, Milione M, Luinetti O, Vindigni C, Chioda C, Albarello L, Bartolini D, Donato F: Comparison of the interobserver reproducibility with different histologic criteria used in celiac disease. Clin Gastroenterol Hepatol. 2007, 5: 838-843. 10.1016/j.cgh.2007.03.019.

    Article  PubMed  Google Scholar 

  6. Ludvigsson J, Lindberg E, Stenhammar L: Tunntarmsbiopsi används för att ställa diagnos. Barnläkaren. 2008, 17-18.

    Google Scholar 

  7. Ludvigsson J, Sjoberg K, Ekbom A: Handläggning av vuxna med misstänkt celiaki – en webbundersökning. Gastrokuriren. 2008, 9-11.

    Google Scholar 

  8. Ivarsson A, Hernell O, Stenlund H, Persson LA: Breast-feeding protects against celiac disease. Am J Clin Nutr. 2002, 75: 914-921.

    CAS  PubMed  Google Scholar 

  9. Dickson BC, Streutker CJ, Chetty R: Coeliac disease: an update for pathologists. J Clin Pathol. 2006, 59: 1008-1016. 10.1136/jcp.2005.035345.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  10. Stenhammar L, Ascher H, Danielsson L, Dannaeus A, Hernell O, Ivarsson A, Lindberg E, Lindquist B, Nivenius K: Small bowel biopsy in Swedish paediatric clinics. Acta Paediatr. 2002, 91: 1126-1129. 10.1080/080352502760311656.

    Article  CAS  PubMed  Google Scholar 

  11. Stenhammar L, Hogberg L, Danielsson L, Ascher H, Dannaeus A, Hernell O, Ivarsson A, Lindberg E, Lindquist B, Nivenius K: How do Swedish paediatric clinics diagnose coeliac disease? Results of a nationwide questionnaire study. Acta Paediatr. 2006, 95: 1495-1497. 10.1080/08035250600636552.

    Article  PubMed  Google Scholar 

  12. Asch DA, Jedrziewski MK, Christakis NA: Response rates to mail surveys published in medical journals. J Clin Epidemiol. 1997, 50: 1129-1136. 10.1016/S0895-4356(97)00126-1.

    Article  CAS  PubMed  Google Scholar 

  13. Sandberg-Bennich S, Dahlquist G, Kallen B: Coeliac disease is associated with intrauterine growth and neonatal infections. Acta Paediatr. 2002, 91: 30-33. 10.1080/080352502753457905.

    Article  CAS  PubMed  Google Scholar 

  14. Askling J, Linet M, Gridley G, Halstensen TS, Ekstrom K, Ekbom A: Cancer incidence in a population-based cohort of individuals hospitalized with celiac disease or dermatitis herpetiformis. Gastroenterology. 2002, 123: 1428-1435. 10.1053/gast.2002.36585.

    Article  PubMed  Google Scholar 

  15. Ludvigsson JF, Ludvigsson J, Ekbom A, Montgomery SM: Celiac Disease and Risk of Subsequent Type 1 Diabetes: A general population cohort study of children and adolescents. Diabetes Care. 2006, 29: 2483-2488. 10.2337/dc06-0794.

    Article  PubMed  Google Scholar 

  16. Ludvigsson JF, Reutfors J, Osby U, Ekbom A, Montgomery SM: Coeliac disease and risk of mood disorders–a general population-based cohort study. J Affect Disord. 2007, 99: 117-126. 10.1016/j.jad.2006.08.032.

    Article  PubMed  Google Scholar 

  17. Troncone R, Ivarsson A, Szajewska H, Mearin ML: Review article: future research on coeliac disease – a position report from the European multistakeholder platform on coeliac disease (CDEUSSA). Aliment Pharmacol Ther. 2008, 27: 1030-1043. 10.1111/j.1365-2036.2008.03668.x.

    Article  CAS  PubMed  Google Scholar 

  18. Mahadeva S, Wyatt JI, Howdle PD: Is a raised intraepithelial lymphocyte count with normal duodenal villous architecture clinically relevant?. J Clin Pathol. 2002, 55: 424-428.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. Brown I, Mino-Kenudson M, Deshpande V, Lauwers GY: Intraepithelial lymphocytosis in architecturally preserved proximal small intestinal mucosa: an increasing diagnostic problem with a wide differential diagnosis. Arch Pathol Lab Med. 2006, 130: 1020-1025.

    PubMed  Google Scholar 

  20. National Institutes of Health Consensus Development Conference Statement on Celiac Disease, June 28–30, 2004. Gastroenterology. 2005, 128: S1-9. 10.1053/j.gastro.2005.02.007.

  21. Rostom A, Murray JA, Kagnoff MF: American Gastroenterological Association (AGA) Institute technical review on the diagnosis and management of celiac disease. Gastroenterology. 2006, 131: 1981-2002. 10.1053/j.gastro.2006.10.004.

    Article  PubMed  Google Scholar 

  22. Korponay-Szabo IR, Szabados K, Pusztai J, Uhrin K, Ludmany E, Nemes E, Kaukinen K, Kapitany A, Koskinen L, Sipka S, et al: Population screening for coeliac disease in primary care by district nurses using a rapid antibody test: diagnostic accuracy and feasibility study. Bmj. 2007, 335: 1244-1247. 10.1136/bmj.39405.472975.80.

    Article  PubMed  PubMed Central  Google Scholar 

  23. Fasano A, Berti I, Gerarduzzi T, Not T, Colletti RB, Drago S, Elitsur Y, Green PH, Guandalini S, Hill ID, et al: Prevalence of celiac disease in at-risk and not-at-risk groups in the United States: a large multicenter study. Arch Intern Med. 2003, 163: 286-292. 10.1001/archinte.163.3.286.

    Article  PubMed  Google Scholar 

  24. Dube C, Rostom A, Sy R, Cranney A, Saloojee N, Garritty C, Sampson M, Zhang L, Yazdi F, Mamaladze V, et al: The prevalence of celiac disease in average-risk and at-risk Western European populations: a systematic review. Gastroenterology. 2005, 128: S57-67. 10.1053/j.gastro.2005.02.014.

    Article  PubMed  Google Scholar 

  25. Hopper AD, Cross SS, Hurlstone DP, McAlindon ME, Lobo AJ, Hadjivassiliou M, Sloan ME, Dixon S, Sanders DS: Pre-endoscopy serological testing for coeliac disease: evaluation of a clinical decision tool. Bmj. 2007, 334: 729-10.1136/bmj.39133.668681.BE.

    Article  PubMed  PubMed Central  Google Scholar 

  26. Hadithi M, von Blomberg BM, Crusius JB, Bloemena E, Kostense PJ, Meijer JW, Mulder CJ, Stehouwer CD, Pena AS: Accuracy of serologic tests and HLA-DQ typing for diagnosing celiac disease. Ann Intern Med. 2007, 147: 294-302.

    Article  PubMed  Google Scholar 

  27. Ludvigsson JF, Ansved P, Falth-Magnusson K, Hammersjo JA, Johansson C, Edvardsson S, Ljungkrantz M, Stenhammar L, Ludvigsson J: Symptoms and Signs Have Changed in Swedish Children With Coeliac Disease. J Pediatr Gastroenterol Nutr. 2004, 38: 181-186. 10.1097/00005176-200402000-00015.

    Article  PubMed  Google Scholar 

  28. Bardella MT, Fredella C, Saladino V, Trovato C, Cesana BM, Quatrini M, Prampolini L: Gluten intolerance: gender- and age-related differences in symptoms. Scand J Gastroenterol. 2005, 40: 15-19. 10.1080/00365520410008169.

    Article  PubMed  Google Scholar 

  29. Green PH, Cellier C: Celiac disease. N Engl J Med. 2007, 357: 1731-1743. 10.1056/NEJMra071600.

    Article  CAS  PubMed  Google Scholar 

  30. Fasano A: Clinical presentation of celiac disease in the pediatric population. Gastroenterology. 2005, 128: S68-73. 10.1053/j.gastro.2005.02.015.

    Article  PubMed  Google Scholar 

  31. Murray JA: Celiac disease in patients with an affected member, type 1 diabetes, iron-deficiency, or osteoporosis?. Gastroenterology. 2005, 128: S52-56. 10.1053/j.gastro.2005.02.029.

    Article  PubMed  Google Scholar 

  32. Green PH: The many faces of celiac disease: clinical presentation of celiac disease in the adult population. Gastroenterology. 2005, 128: S74-78. 10.1053/j.gastro.2005.02.016.

    Article  PubMed  Google Scholar 

  33. Ciacci C, Cirillo M, Sollazzo R, Savino G, Sabbatini F, Mazzacca G: Gender and clinical presentation in adult celiac disease. Scand J Gastroenterol. 1995, 30: 1077-1081. 10.3109/00365529509101610.

    Article  CAS  PubMed  Google Scholar 

  34. Anson O, Weizman Z, Zeevi N: Celiac disease: parental knowledge and attitudes of dietary compliance. Pediatrics. 1990, 85: 98-103.

    CAS  PubMed  Google Scholar 

  35. Kumar P: Dietary compliance and coeliac disease. 1998, Tampere: The Finnish Coeliac Society

    Google Scholar 

  36. Hogberg L, Grodzinsky E, Stenhammar L: Better dietary compliance in patients with coeliac disease diagnosed in early childhood. Scand J Gastroenterol. 2003, 38: 751-754. 10.1080/00365520310003318.

    Article  CAS  PubMed  Google Scholar 

  37. Fabiani E, Taccari LM, Ratsch IM, Di Giuseppe S, Coppa GV, Catassi C: Compliance with gluten-free diet in adolescents with screening-detected celiac disease: A 5-year follow-up study. J Pediatr. 2000, 136: 841-843. 10.1067/mpd.2000.105351.

    Article  CAS  PubMed  Google Scholar 

Pre-publication history

Download references

Acknowledgements

This project was supported by a grant from The Swedish Society of Medicine, the Örebro University Hospital, the Swedish Research Council, the Sven Jerring Foundation, the Örebro Society of Medicine, the Karolinska Institutet, the Clas Groschinsky Foundation, the Juhlin Foundation, the Majblomman Foundation and the Swedish Coeliac Society.

We would like to thank Åke Östberg for the photomicrographs of small intestinal biopsies (Figure 1).

We would like to thank the company Equalis for access to data on the accuracy of histopathological accuracy of small intestinal biopsies.

We would like to thank Paul Blomquist for valuable comments on the manuscript.

Data from two of the included web surveys of this paper have previously been published in Swedish-language non-Medline-indexed publications that target Swedish gastroenterologists (publication: "Gastrokuriren") and pediatricians ("Barnläkaren").

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Jonas F Ludvigsson.

Additional information

Competing interests

The authors declare that they have no competing interests.

Authors' contributions

JFL designed the study, wrote the ethics application, collected the data, analyzed the statistics and wrote the manuscript. LB helped JFL manage the data and critically reviewed the manuscript. SM, FG and AE critically reviewed the manuscript and contributed to the study design.

Electronic supplementary material

12876_2008_311_MOESM1_ESM.doc

Additional File 1: Table S1. Comorbidity defined by SnoMed codes in patients with inflammation or villous atrophy (VA) – a single centre evaluation. (DOC 74 KB)

Authors’ original submitted files for images

Below are the links to the authors’ original submitted files for images.

Authors’ original file for figure 1

Authors’ original file for figure 2

Rights and permissions

This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Reprints and permissions

About this article

Cite this article

Ludvigsson, J.F., Brandt, L., Montgomery, S.M. et al. Validation study of villous atrophy and small intestinal inflammation in Swedish biopsy registers. BMC Gastroenterol 9, 19 (2009). https://doi.org/10.1186/1471-230X-9-19

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/1471-230X-9-19

Keywords