Classification chaos in coeliac disease: Does it really matter?

A Deep Learning-Based Approach to Estimate Paneth Cell Granule Area in Celiac Disease

CONTEXT.—

Changes in Paneth cell numbers can be associated with chronic inflammatory diseases of the gastrointestinal tract. So far, no consensus has been achieved on the number of Paneth cells and their relevance to celiac disease (CD).

OBJECTIVES.—

To compare crypt and Paneth cell granule areas between patients with CD and those without CD (non-CD) using an artificial intelligence-based solution.

DESIGN.—

Hematoxylin-eosin-stained sections of duodenal biopsies from 349 patients at the McGill University Health Centre were analyzed. Of these, 185 had a history of CD and 164 were controls. Slides were digitized, and NoCodeSeg, a code-free workflow using open-source software (QuPath, DeepMIB), was implemented to train deep learning models to segment crypts and Paneth cell granules. The total area of the entire analyzed tissue, epithelium, crypts, and Paneth cell granules was documented for all slides, and comparisons were performed.

RESULTS.—

A mean intersection-over-union score of 88.76% and 91.30% was achieved for crypt areas and Paneth cell granule segmentations, respectively. On normalization to total tissue area, the crypt to total tissue area in CD was increased and the Paneth cell granule area to total tissue area decreased when compared to non-CD controls.

CONCLUSIONS.—

Crypt hyperplasia was confirmed in CD compared to non-CD controls. The area of Paneth cell granules, an indirect measure of Paneth cell function, decreased with increasing severity of CD. More importantly, our study analyzed complete hematoxylin-eosin slide sections using an efficient and easy to use coding-free artificial intelligence workflow.

Navigating the Challenges of Gluten Enteropathy and Infertility: The Role of Celiac-Related Antibodies and Dietary Changes

Celiac disease (CD) is an autoimmune condition that is initiated in genetically susceptible individuals by the exposure of the intestines to gluten, and the early start of symptoms is related to malabsorption. Atypical variants of the illness are often identified in adulthood and are frequently associated with manifestations outside of the intestines, including metabolic osteopathy, anemia, and dermatitis herpetiformis. But also, empirical data suggest a correlation between CD and reproductive abnormalities, including repeated abortions. Infertility and repeated miscarriages frequently manifest in women diagnosed with CD and may serve as the initial clinical indication of a subclinical form. Furthermore, the condition may manifest as amenorrhea, infertility, and the delivery of infants with a low birth weight. Regarding the mechanisms of CD in infertility, along with the anti-tTG action to hinder the invasiveness of trophoblast, these antibodies could damage endometrial angiogenesis, which has been shown in in vitro models with human endometrial cells and in vivo in murine models. Another important aspect is the role of nutrient deficiencies, such as zinc deficiency (connected to impaired hormone production, secondary amenorrhea, and pre-eclampsia) and folic acid, etc. Therefore, our objective was to conduct a comprehensive review of the existing literature pertaining to this specific topic and to elucidate the role of the autoantibodies in its pathogenesis.

Celiac Disease: Updates on Pathology and Differential Diagnosis

Celiac disease is a gluten-triggered immune-mediated disorder, characterized by inflammation of the enteric mucosa following lymphocytic infiltration and eventually resulting in villous blunting. There have been many developments in refining diagnostic laboratory tests for celiac disease in the last decade. Biopsy-sparing diagnostic guidelines have been proposed and validated in a few recent prospective studies. However, despite these developments, histologic evaluation of duodenal mucosa remains one of the most essential diagnostic tools as it helps in the diagnosis of celiac disease in individuals who do not fulfill the biopsy-sparing diagnostic criteria and in those not responding to a gluten-free diet. Histologic evaluation also allows for the assessment of mucosal recovery after treatment and in the identification of concurrent intestinal diseases. Therefore, pathologists should be familiar with the histologic spectrum of celiac disease and need to be aware of other disorders with similar symptoms and histopathology that may mimic celiac disease. This review aims to provide pathologists with updates on celiac laboratory testing, biopsy-sparing diagnostic criteria, histopathology, complications, and differential diagnoses of celiac disease.

Diagnosing celiac disease: A critical overview

The diagnosis of celiac disease (CD) no longer rests on a malabsorptive state or severe mucosal lesions. For the present, diagnosis will always require the gold-standard of a biopsy, interpreted through its progressive phases (Marsh classification). Marsh classification articulated the immunopathological spectrum of gluten-induced mucosal changes in association with the recognition of innate (Marsh I infiltration) and T cell-based adaptive (Marsh II, and the surface re-organisation typifying Marsh III lesions) responses. Through the Marsh classification the diagnostic goalposts were considerably widened thus, over its time-course, permitting countless patients to begin a gluten-free diet but who, on previous criteria, would have been denied such vital treatment. The revisions of this classification failed to provide additional insight in the interpretation of mucosal pathology. Morever, the subclassification of Marsh 3 imposed an enormous amount of extra work on pathologists with no aid in diagnosis, treatment, or prognosis. Therefore, it should now be apparent that if gastroenterologists ignore these sub-classifications in clinical decision-making, then on that basis alone, there is no need whatsoever for pathologists to persist in reporting them. Since new treatments are under critical assessment, we might have to consider use of some other higher level histological techniques sensitive enough to detect the changes sought. A promising alternative would be to hear more voices from imaginative histopathologists or morphologists together with some more insightful approaches, involving molecular-based techniques and stem cell research may be to evaluate mucosal pathology in CD.

Discriminant value of IEL counts and distribution pattern through the spectrum of gluten sensitivity: a simple diagnostic approach

Intraepithelial lymphocytosis (IELosis) with or without villous abnormality is a characteristic feature of gluten sensitivity (GS) including celiac disease (CD) and non-celiac-GS, although various conditions may also be associated with IELosis. In order to distinguish GS from the other causes of IELosis, a threshold for IEL counts is necessary. We aimed to determine a cut-off value for IELs and monitor its value in the spectrum of GS in a large cohort. For this purpose, the duodenal biopsies from four groups of individuals including Types 1 (n = 88) and 3 (n = 92) CD, non-CD IELosis (n = 112), and control (n = 82) cases, all strictly defined by their clinical, laboratory, and serologic features, were evaluated. The number of IELs/100 enterocytes and their distribution pattern on H&E- and CD3-immunostained sections were assessed for each group. Kruskal-Wallis test and ROC curve analysis for discriminant value were employed for statistics. The IEL counts showed an increasing trend through the spectrum of mucosal pathology including controls (12.06; 21.40), non-CD IELosis (28.62; 39.46), Type 1 CD (49.27; 60.15), and Type 3 CD (58.53; 71.74) both on H&E- and CD3-immunostained sections, respectively (p < 0.001). ROC analysis revealed 20.5 on H&E and 28.5 on CD3 as the IEL cut-off values with a sensitivity of 95.9 and 87.7% and a specificity of 98.8% and 93.9%, respectively, for controls. IELs showed a diffuse distribution pattern per biopsy piece and per villus (90.9%, 100%, respectively) in nearly all of Type 1 CD cases (p < 0.001). An IEL cut-off value of 20.5 on H&E together with a diffuse distribution pattern seem to be the most discriminant features for the diagnosis of CD, even for the milder forms of the disease.