γδ+ T-Cells Is a Useful Biomarker for the Differential Diagnosis between Celiac Disease and Non-Celiac Gluten Sensitivity in Patients under Gluten Free Diet

Abstract

Background: The differential diagnosis between patients with celiac disease (CD) and non-celiac gluten sensitivity (NCGS) is difficult when a gluten-free diet (GFD) has been initiated before the diagnostic work-up. Isolated increases in TCRγδ+ and celiac lymphogram (increased TCRγδ+ plus decreased CD3-) may enable differential diagnosis in this challenging clinical setting. This study evaluated: (1) the accuracy of %TCRγδ+ and celiac lymphogram for diagnosing CD before and after GFD and for differentiation with NCGS; (2) TCRγδ+ kinetics at baseline and after starting GFD in both CD and NCGS.

Methods: The inclusion criteria were patients with CD (n = 104), NCGS (n = 37), and healthy volunteers (n = 18). An intestinal biopsy for intraepithelial lymphogram by flow cytometry was performed at baseline and after GFD. The optimal cutoff for CD diagnostic accuracy was established by maximizing the Youden index and via logistic regression.

Results: %TCRγδ+ showed better diagnostic accuracy than celiac lymphogram for identifying CD before and after GFD initiation. With a cutoff > 13.31, the accuracy for diagnosing CD in patients under GFD was 0.88 [0.80-0.93], whereas the accuracy for diagnosing NCGS (%TCRγδ+ ≤ 13.31) was 0.84 [0.76-0.89]. The percentage of TCRγδ+ cells showed differential kinetics between CD (baseline 22.7% [IQR, 16.4-33.6] vs. after GFD 26.4% [IQR, 17.8-36.8]; p = 0.026) and NCGS (baseline 9.4% [IQR, 4.1-14.6] vs. after GFD 6.4% [IQR, 3.2-11]; p = 0.022).

Conclusion: TCRγδ+ T cell assessment accurately diagnoses CD before and after a GFD. Increased TCRγδ+ was maintained in the long term after GFD in CD but not in NCGS. Altogether, this suggests the potential usefulness of this marker for the differential diagnosis of these two entities in patients on a GFD.

Keywords: celiac disease; flow cytometry; gluten-free diet; intraepithelial lymphogram; non-celiac gluten sensitivity; γδ T cells.

Figure 1

Study flow chart. Abbreviations: HUMT: Hospital Universitari Mútua Terrassa; GFD, gluten-free diet; IBS: irritable bowel syndrome; NSAID: nonsteroidal anti-inflammatory drug; CD, celiac disease; NCGS, non-celiac gluten sensitivity.

Figure 2

ROC curve to assess the accuracy of the TCRγδ+ IEL for identifying patients with celiac disease (gold standard healthy individuals). Abbreviations: AUC: area under the curve; ROC: receiver operating characteristic; TCR: T-cell receptor.

Figure 3

Median with interquartile range and dot and line diagrams describing the evolution of TCRγδ+ (green lines) and CD3− cells (blue lines) before and after a gluten-free diet (GFD) in patients with celiac Marsh 3, celiac Marsh 1, and non-celiac gluten sensitivity (NCGS). Black dotted line indicates the cutoff point for TCRγδ+ cells at 13.31%. Red lines indicate patients with nonpersistence of increased γδ T-subsets after a GFD (final TCRγδ+ cells ≤ 13.31%). Abbreviations: TCR: T-cell receptor; IEL: intraepithelial lymphocyte.

Figure 4

Baseline and final distributions of the percentage of TCRγδ+ cells (A,C) and percentage of CD3− cells (B,D) in the three groups: Marsh 0 healthy individuals, Marsh 0 non-celiac gluten sensitivity patients and Marsh 1–3 celiac disease patients. Abbreviations: TCR: T-cell receptor; IEL: intraepithelial lymphocyte.

Figure 5

Pathological and intestinal cytometry images illustrating one case of three following categories in the study: the first row shows a patient with Marsh 3 celiac disease (A–D), the second row a patient with Marsh 1 celiac disease (E–H), and the third row a patient with non-celiac gluten sensitivity (I–L). The first two columns (A,B,E,F,I,J) show the images of patients on a gluten-containing diet, and the third and fourth columns (C,D,G,H,K,L) show the images of patients on a gluten-free diet. (A) Immunohistochemical staining of CD3 lymphocytes in duodenal biopsies with atrophy (Marsh 3) (orig. mag. x20). The red arrows indicate the accumulation of CD3 lymphocytes in the atrophic mucosa. (B) Intestinal cytometry panel showing increased γδ T-subsets (red circles) with GCD in a sample of duodenal atrophy (Marsh 3). (C) Immunohistochemical staining of CD3 lymphocytes in the recovered duodenal mucosa (Marsh 0) with a GFD (original magnification x20). The red arrows indicate the decrease of CD3 lymphocytes in the duodenal villi (orig. mag. x20). (D) Intestinal cytometry panel showing persistent increase of γδ T-subsets (red circles) with a GFD in a sample of recovered duodenal mucosa (Marsh 0). (E) Immunohistochemical staining of CD3 lymphocytes in duodenal biopsies with lymphocytic enteritis (Marsh 1) (orig. mag. x20). The red arrows indicate the accumulation of CD3 lymphocytes in the duodenal mucosa. (F) Intestinal cytometry panel showing in-creased γδ T-subsets (red circles) with GCD in a sample of lymphocytic enteritis (Marsh 1). (G) Immunohistochemical staining of CD3 lymphocytes in the recovered duodenal mucosa (Marsh 0) with a GFD (orig. mag.x20). The red arrows indicate the decrease of CD3 lymphocytes in the duodenal villi. (H) Intestinal cytometry panel showing persistent in-crease of γδ T-subsets (red circles) with a GFD in a sample of recovered duodenal mucosa (Marsh 0). (I) H&E staining of normal duodenal biopsies from a patient with NCGS (orig. mag. x4). (J) Intestinal cytometry panel showing low values of γδ T-subsets with a GCD in a sample of normal duodenal mucosa (Marsh 0) from a patient with NCGS. (K) H&E staining showing persistence of normal duodenal biopsies from a patient with NCGS (orig. mag. x4). (L) Intestinal cytometry panel showing persistent low values of γδ T-subsets with a GFD in a sample of normal duodenal mucosa (Marsh 0) from a patient with NCGS. Abbreviations: GFD: Gluten-free diet; GCD: Gluten-containing diet; H&E: Hematoxylin and eosin; orig. mag.: original magnification; NCGS: Non-celiac gluten sensitivity.