Diagnostic yield of duodenal biopsies embedded in paraffin blocks without luminal orientation in celiac disease

Authors

Abstract

AimIn histomorphologic evaluation of celiac disease, to provide the best possible representation of surface epithelium and villi, the luminal side of small intestinal biopsy samples should be identified and embedded in paraffin blocks accordingly.In this study, we aim to evaluate representation of surface epithelium and villi, and scorability of non-oriented duodenal biopsies.
MethodsA total of 1022 endoscopic biopsy samples taken from 372 locations (duodenal bulb and second-part for each patient) of 186 patients were re-evaluated based on the representation of the surface epithelium and villi.Biopsy samples containing et least ‘discontinuous epithelial segments of more than 100 cells’ with at least ‘3-4 adjacent villi' were considered scorable.Scorabilities were determined based on the most scorable biopsy sample for locations and the most scorable location for the patients.
ResultsThe scorability rate was 60.37% for the biopsy samples, 74.19% for the duodenal bulb biopsies, 85.49% for the second-part, and 94.62% for the patients.The scorability rate was significantly higher in patients aged over 18 years, in patients with tTG IgA values below 200 RU/mL, and in second-part biopsies (p<0.05).
ConclusionConsidering that endoscopy is an invasive procedure, the scorability of non-oriented duodenal biopsy samples is low.However, even if no luminal orientation method can be used, the scorability of patients can be significantly increased by increasing the number of biopsy samples and biopsy locations taken from the patient.

Keywords

Introduction

Celiac disease (CD) is an immune-mediated disorder triggered by gluten.^1,2,3CD primarily affects the small intestine.Patients with classic CD typically experience gastrointestinal symptoms and signs of malabsorption upon gluten consumption.^1,3 Small intestinal biopsy is generally performed for diagnosis, except in selected pediatric patients.^1,4However, histomorphologic changes observed in biopsy samples should be interpreted alongside clinical symptoms, signs of malabsorption, and serologic and genetic tests.⁴
In 1992, the Marsh scoring system was developed to categorize small intestinal biopsies histomorphologically according to the degrees of intraepithelial lymphocytosis, villus atrophy, and crypt hyperplasia.Hematoxylin and eosin (H&E)-stained sections are usually sufficient for scoring.2In most biopsies, evaluating crypt morphology is not problematic.However, scoring cannot be reliably performed when the surface epithelium needed to assess intraepithelial lymphocytosis or the villi are inadequately represented.In this context, it is recommended that the luminal side of biopsy samples be identified by the pathologist/pathology technician and be embedded in paraffin blocks accordingly.^2,5,6,7
In this study, we aim to evaluate the representation of surface epithelium and villi, and the scorability of non-oriented duodenal biopsies in CD prediagnosis.

Materials and Methods

Two hundred eighty patients who underwent endoscopic duodenal biopsy with a prediagnosis of CD between 2021 and 2024 at Şanlıurfa Training and Research Hospital, were determined by scanning the pathology report archive.Age, sex, epicrisis report, endoscopy report, and blood tissue transglutaminase-immunoglobulin A antibody (tTG IgA) test values (reference: 0-20 RU/mL) of the patients were obtained from the hospital system.Six patients who were followed up with a gluten-free diet, four with IgA deficiency, and 84 patients for whom tTG IgA values not available at the time of the biopsy (within the period covering two months before the biopsy date) were excluded from the study.
Biopsy samples were taken from the duodenal bulb and second-part of all patients and sent to the pathology laboratory in separate bottles.All samples that were sent in the same bottle were embedded in the same paraffin block randomly without any luminal orientation method by pathology technicians.Two serial H&E-stained sections were obtained from each paraffin block.These first two serial H&E-stained sections from each location of the patients, prepared during routine pathologic examinations, were obtained from the pathology section archive for re-evaluation.The biopsy samples on the first two serial H&E-stained sections were individually numbered so that the same pieces received the same number.They were evaluated by one pathologist for the representation of the surface epithelium and villi under a light microscope (Olympus, BX43).To ensure internal control, two different H&E-stained sections belonging to the same paraffin block were evaluated separately at different times and the evaluation results were compared.In the comparison, for the analysis, a low evaluation of these parameters was considered significant.
The evaluation of surface epithelium was classified as follows: E0: Unrepresented, E1: Discontinuous epithelial segments of less than 100 cells, E2: Discontinuous epithelial segments of more than 100 cells, E3: Continuous epithelium (Figure 1).The evaluation of villi was classified as follows: V0: Unrepresented, V1: 1-2 adjacent villi, V2: 3-4 adjacent villi, V3: Many adjacent villi or completely atrophic flat surface epithelium (Figure 2).The evaluation of biopsy samples scorability was based on the classification of surface epithelium and villi. Biopsy samples containing et least ‘discontinuous epithelial segments of more than 100 cells’ (E2-3) with at least ‘3-4 adjacent villi' (V2-3) were considered scorable (S1). The scorability of biopsy samples, biopsy locations (determined based on the most scorable sample for each location), and patients (determined based on the most scorable location for each patient) were established.
Ethical ApprovalThis study was approved by the Ethics Committee of Harran University (Date: 12.02.2024, Decision No: HRÜ/24.01.59).
Statistical AnalysisStatistical analyses were performed using SPSS version 26.0 (IBM Corp., Armonk, NY, USA). Categorical variables were expressed as number and percentage. Associations between categorical variables were analyzed using the χ² test or Fisher exact test, as appropriate. A 2-sided p value <0.05 was considered statistically significant.
Reporting GuidelinesThis study was reported in accordance with the STROBE guideline.

Results

A total of 1022 endoscopic biopsy samples taken from 372 locations (bulb and second-part of the duodenum for each patient) of 186 patients were evaluated.
Clinical, laboratory, endoscopic, and histomorphologic findings were given in Table I.Most of the patients were female (65.59%), and under 18 years old (96.77%).Their tTG IgA values were mostly over 200 RU/mL (53.23%).They had frequently abnormal endoscopic findings.Mostly biopsies from the locations were multiple.Surface epithelium were represented frequently as discontinuous epithelial segments of more than 100 cells (56.26%).Villi were represented frequently as 3-4 adjacent villi (40.80%).
The scorability rate was 60.37% for the biopsy samples, 74.19% for the duodenal bulb biopsies, 85.49% for the second-part, and 94.62% for the patients.
The relationships between clinical, laboratory, and endoscopic findings with histomorphologic findings, and scorability were given in Table II.The scorability rate was significantly higher in patients aged over 18 years, in patients with tTG IgA values below 200 RU/mL, and in second-part biopsies (p<0.05).

Discussion

Even if CD persists in adulthood, requiring lifelong glutenfree diet and control, the diagnosis is generally made in pediatric clinics.^1,2,3Pediatric developmental periods are classified according to ages 0-2 as infancy, 2-6 as early childhood, 6-12 as middle childhood, and 12-18 as adolescence (late childhood).⁸The period when patients are frequently diagnosed is infancy, when gluten is initially included in the diet.² However, most of our study patients were in middle childhood (40.86%).Considering that the study was conducted in the city of Türkiye with the highest birth rate and the second lowest average education period according to 2022 data, this situation may be related to late admission to health centers due to the low sociocultural level of the study region.^9,10
Female predominance in CD has been reported in the literature.¹In this study, all patients underwent duodenal biopsy via CD prediagnosis, but histomorphologic evaluations or definitive diagnoses determined according biopsy results were not taken into account in patient selection.The majority of the patients were female (65.59%).
tTG IgA is the first recommended serologic test to be performed in the prediagnosis of CD.1,2tTG IgA should be considered significant when it increases at least 5-10 times in patients who have a diet containing gluten and do not have low total IgA values.^1,2,3,6The tTG IgA values were markedly elevated in most of the study patients; exceeding 5 times the normal limit (>100 RU/mL) in 67.75% and exceeding 10 times the normal limit (>200 RU/mL) in 53.23%.It is important to note that seronegativity in CD has been reported in the literature, albeit at a low rate (1.03-4.16%).¹¹Although some of our study patients showed no significant seropositivity, physicians taking endoscopic biopsies from these patients may indicate physicians awareness of this issue.
CD affects the small intestine with decreasing rates from the duodenum to the ileum.^1,4,7Therefore, the duodenum is routinely examined in patients with a CD prediagnosis.Endoscopic procedures include examination of the proximal two parts of the duodenum; the first part (bulb) and second-part.¹²Frequently, endoscopic abnormalities were described in the study patients, mostly nonspecific findings such as hyperemia, nodulation, or ulceration in the bulb (78.49%) and suspicious findings for CD such as scalloping or cracked mud appearance in the second-part (82.80%).[4.5]Scalloping and cracked mud appearance are sensitive but not specific enough in adults and sensitivities decrease in pediatric patients.^1,6
CD affects the duodenum in a patchy manner, so multiple biopsies are advised—typically 2 from the bulb and 4 from the second-part.^1,2,4,6Unfortunately, in 2011 Lebwohl et al.reported that four or more biopsy samples were taken in only 39% of patients who underwent biopsy with a CD prediagnosis in 43 states across the United States.¹³In our study the rates were higher, and 84.95% of the bulb biopsies and 93.01% of the second-part biopsies contained multiple samples.As recommended, the most frequently taken number of samples from the bulb was two (73.12%), from the second-part it was four (60.22%).This increase in the tendency to perform multiple biopsies may be related to the awareness that has come with the increase in studies supporting the diagnostic importance of taking multiple biopsies over the years.
For optimal scoring, it is recommended to evaluate intraepithelial lymphocytosis on 100 epithelial cells and villous morphology on at least 3-4 adjacent villi.^4,7To provide the best possible representation of these luminal structures, the luminal side of biopsy samples should be identified and embedded in paraffin blocks accordingly.^2,5,6,7Biopsy orientation can be established by examining the villi with a hand lens and dissection microscope.⁴However, in routine practice, it takes time and requires experience.¹⁴Also, it is too difficult for pathologists/pathology technicians to distinguish the luminal side of biopsy samples, which are sent to the pathology laboratory in formalin-filled biopsy bottles and shrunken and hardened after fixation.Some authors have recommended that endoscopists use acetate cellulose filters.^6,8,15These filters allow endoscopists to separate biopsies according to their locations, line them up from proximal to distal, and indicate the luminal surface.This allows the pathologist/pathology technician to embed biopsies in paraffin with the correct surface orientation quickly.^8,15However, it increases financial burden and has been observed that acetate cellulose filters do not contribute significantly to biopsy scorability.⁶
In this study, biopsy samples were randomly embedded in paraffin blocks by pathology technicians without using any orientation method.In most of the samples, discontinuous epithelial segments of more than 100 cells (56.26%), and 3-4 adjacent villi (40.80%) were observed.Most of the samples were scorable (60.37%).
In a multicentre study from Europe, only 10.7% of biopsies had been evaluated as unscorable due to poor orientation.However, whether any orientation method was used was not specitied in this study.16On the other hand, the rate of our non-oriented biopsy samples not suitable for scoring was 39.63%.Considering that endoscopy is an invasive procedure, this rate was quite high.When all samples of the same location were evaluated together, the scorability rate of the bulb-localized biopsies was 74.19%, the scorability rate of the second-part-localized biopsies was 85.49%.When both locations were evaluated together, the scorability rate of the patients was 94.62%.As a result, when 1022 samples were examined, 39.63% were unscorable, but when the number of locations and samples were increased, the rate of unscorability of the 186 patients decreased to 5.38%.Pais et al.reported that two samples were adequate for diagnosis of CD with a rate of 90%, three samples with a rate of 95%, and four samples with a rate of 100%.17In another study with a large population, the rates of CD diagnoses were found significantly higher in patients from whom more than four samples were taken.1All these findings support the need to increase the number of biopsies to increase scorability.
The representation statuses of the surface epithelium (p=0.840), villi (p=0.770), and scorability (p=0.510) of the biopsy samples were similar in females and males in our study.This situation may exclude differences that might arise from anatomic features of sexes.However, the results are not statistically significant.
In pediatric patients, verbal communication is more difficult, which worsens procedure compatibility in these patients.The procedure area is narrower, so special skills and equipment are needed.Smaller samples can be obtained with pediatric endoscopic forceps.18However, Dandalides et al.compared standard and jumbo forceps in adults and found that biopsies taken using standard forceps were sufficient in all patients and they argued that the biopsy diameter was not important.¹⁶In patients aged over 18 years, representation of continuous surface epithelium (p=0.010), representation of many adjacent villi or complately atrofic flat villi (p<0.001), and scorability (p= 0.010) were significantly higher.These increases in rates might have occurred due to the decrease in these negativities in adults.
In a multicenter study, tTG IgA was found to be 98% sensitive and 75% specific for detecting villous atrophy.20A correlation between increasing tTG IgA values and the severity of villous atrophy has been reported.¹⁷ In our study, we observed a decrease in the representation of continuous surface epithelium (p<0.001), many adjacent villi or complately atrofic flat villi (p<0.001) as tTG IgA values increased.Additionally, patients with tTG IgA values exceeding 200 RU/mL exhibited lower scorability compared with other patients (p<0.001).Given that increasing tTG IgA values indirectly indicate the presence and severity of villous atrophy, this situation may be attributed to tissue quality deterioration in the affected area or difficulties in obtaining biopsy samples at the correct vertical angle due to mucosal fold deterioration.
Scalloping and cracked mud appearance are usually sensitive for CD.1In the study of Ravelli et al., in 80 pediatric patients who had villous atrophy in pathologic examinations, cracked mud appearance was seen in 100% and scalloping in 70% in endoscopic examinations.{22]In most of our patients whose endoscopy was reported as normal, the surface epithelium was represented as continuous.In patients with any abnormalities that were described endoscopically, the surface epithelium was mostly represented as discontinuous segments of more than 100 cells (p<0.001).In most patients, 3-4 adjacent villi were represented, regardless of endoscopic findings.However, the representation rate of many adjacent villi was higher in patients whose endoscopy was described as normal, and the rate of representation of 3-4 adjacent villi was higher in patients whose endoscopy was described as abnormal (p=0.030).Given that endocopic abnormalities (expecially scalloping and cracked mud appearance) indirectly indicate the presence of villous atrophy, this situation may be attributed to tissue quality deterioration in the affected area or to difficulties in obtaining biopsy samples at the correct vertical angle due to mucosal fold deterioration as tTG IgA values.
In samples taken from both the bulb or second-part, the surface epithelium was frequently represented as discontinuous segments of more than 100 cells.However, continuous surface epithelium was more common in the second-part (p=0.0170).In samples taken from both locations, mostly 3-4 adjacent villi were represented.Additionally, the rates of samples with no villi or 1-2 adjecent villi representation were higher in the bulb, and the rates of samples with 3-4 adjecent villi or many adjacent villi or completely atrophic flat villi were higher in the second-part (p<0.001).Scorability (p<0.001) were higher in the second-part.In a systematic review and meta-analysis analyzing 16 studies, the diagnostic yields of bulbs and second-part biopsies were found to be similar in adults.22However, in another study, the adequacy of bulb biopsies was lower in pediatric patients.6Considering that most of our patients were pediatric, this study supports our findings.These differences between the bulb and second-part may be related to difficulties in processing due to the anatomy of the locations and the inability to adjust the correct vertical angle due to these difficulties.Additionally, it may be more pronounced in the pediatric population because of poorer patient compliance, the narrower procedure area, and special skills and equipment required for the procedures.

Limitations

This study was conducted at a single center and included predominantly pediatric patients. Histomorphologic evaluations were performed by a single pathologist, and interobserver variability was not assessed. The retrospective design may limit the generalizability of the findings.

Conclusion

Considering that endoscopy is an invasive procedure, the scorability of non-oriented duodenal biopsy samples is low.Even if no luminal orientation method can be used, the scorability of patients can be significantly increased by increasing the number of biopsy samples and biopsy locations taken from patients.Therefore, especially when orientation is not possible due to reasons such as workload, lack of equipment, and inexperience, multiple biopsy samples from different parts of the duodenum (such as bulb and second-part) should be taken, when there is a prediagnosis of CD.

Declarations

Abbreviations

CD: Celiac disease
H&E: Hematoxylin and eosin
SPSS: Statistical package for the social sciences
STROBE: Strengthening the reporting of observational studies in epidemiology
tTG IgA: Tissue transglutaminase immunoglobulin A antibody

References

1. Kefeli A. Celiac disease. Turk J Curr Gastroenterol. 2023;25(1):37-47.
   PubMed Google Scholar
2. Villanacci V, Vanoli A, Leoncini G, et al. Celiac disease: histology-differential diagnosis-complications: a practical approach. Pathologica. 2020;112(3):186-196. doi:10.32074/1591-951x-157
   Article PubMed Google Scholar
3. Sahin Y. Celiac disease in children: a review of the literature. World J Pediatr. 2021;10(4):53-71. doi:10.5409/wjcp.v10.i4.53
   Article PubMed Google Scholar
4. Robert ME, Gibson JA. Inflammatory disorders of the small intestine/disorders of malabsorption/gluten-sensitive enteropathy (celiac disease). In: Odze RD, Goldblum JR, eds. Surgical Pathology of the GI Tract, Liver, Biliary Tract, and Pancreas. Elsevier Saunders; 2015:402-412.
   PubMed Google Scholar
5. Green PHR. Celiac disease: how many biopsies for diagnosis? Gastrointest Endosc. 2008;67(7):1088-1090. doi:10.1016/j.gie.2007.12.035
   Article PubMed Google Scholar
6. Popp A, Balaba VD, Mäki M. Celiac disease. In: Gershman G, Thomson MA, eds. Practical Pediatric Gastrointestinal Endoscopy. Wiley-Blackwell; 2021:207-212. doi:10.1002/9781119423492.ch23
   Article PubMed Google Scholar
7. Dickson BC, Streutker CJ, Chetty R. Coeliac disease: an update for pathologists. J Clin Pathol. 2006;59:1008-1016. doi:10.1136/jcp.2005.035345
   Article PubMed Google Scholar
8. Gürsoy F, Özaslan H. Çocuk gelişimine giriş/gelişim dönemleri. In: Aral N, ed. Çocuk Gelişimi. Anadolu Üniversitesi; 2018:8-11.
   PubMed Google Scholar
9. Kaur J, Jialal I. Hashimoto thyroiditis. StatPearls Publishing. Accessed February 9, 2025. https://www.ncbi.nlm.nih.gov/books/NBK459262
   PubMed Google Scholar
10. Turkish Statistical Institute. Birth statistics, 2022. Accessed May 15, 2024. https://data.tuik.gov.tr/Bulten/Index?p=Birth-Statistics-2022-49673
    PubMed Google Scholar
11. Dhingra K, Maharshi S, Sapra B, Ratra S, Sharma SS, Nijhawan S. Seronegative celiac disease: frequently encountered yet undiagnosed clinical entity. Middle East J Dig Dis. 2021;13(1):35-42. doi:10.34172/mejdd.2021.201
    Article PubMed Google Scholar
12. Peña AS. What is the best histopathological classification for celiac disease? Does it matter? Gastroenterol Hepatol Bed Bench. 2015;8(4):239-243.
    PubMed Google Scholar
13. Lebwohl B, Kapel RC, Neugut AI, Green PH, Genta RM. Adherence to biopsy guidelines increases celiac disease diagnosis. Gastrointest Endosc. 2011;74:103-109. doi:10.1016/j.gie.2011.03.1236
    Article PubMed Google Scholar
14. Ravelli A, Villanacci V. Tricks of the trade: how to avoid histological pitfalls in celiac disease. Pathol Res Pract. 2012;208:192-202. doi:10.1016/j.prp.2012.01.008
    Article PubMed Google Scholar
15. Villanacci V, Del Sordo R, Casella G, et al. The correct methodological approach to the diagnosis of celiac disease: the point of view of the pathologist. Gastroenterol Hepatol Bed Bench. 2023;16(2):129-135.
    PubMed Google Scholar
16. Dandalides SM, Carey WD, Petras R, Achkar E. Endoscopic small bowel mucosal biopsy: a controlled trial evaluating forceps size and biopsy location in the diagnosis of normal and abnormal mucosal architecture. Gastrointest Endosc. 1989;35(3):197-200. doi:10.1016/s0016-5107(89)72757-7
    Article PubMed Google Scholar
17. Ierardi E, Losurdo G, Piscitelli D, et al. Seronegative celiac disease: where is the specific setting? Gastroenterol Hepatol Bed Bench. 2015;8(2):110-116.
    PubMed Google Scholar

Additional Information

Publisher’s Note
Bayrakol MP remains neutral with regard to jurisdictional and institutional claims.

Rights and Permissions

Creative Commons License

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC 4.0). To view a copy of the license, visit https://creativecommons.org/licenses/by-nc/4.0/

View full license details →

About This Article

Received:

May 11, 2026

Accepted:

June 15, 2026

Published Online:

June 19, 2026