Immunohistochemical Stains for CD3 and CD8 Do Not Improve Detection of Gluten Sensitivity in Duodenal Mucosal Biopsies
Rachel M Hudacko, Xi K Zhou, Rhonda K Yantiss. Weill Medical College of Cornell University, New York, NY
Background: Patients with gluten sensitivity may have duodenal mucosal biopsies with preserved villous architecture and increased numbers of intraepithelial lymphocytes (IELs). Some authors advocate use of CD8 and/or CD3 immunostains to improve IEL detection, resulting in a growing trend to perform these stains “up front” on all duodenal biopsies. However, the added value of immunohistochemistry in detecting increased IELs when H&E stains are interpreted to be normal has not been systematically evaluated. The purpose of this study was to determine whether CD3 and CD8 stains improve detection of gluten sensitivity in duodenal samples compared to H&E stains alone.
Design: This retrospective study included 192 duodenal biopsies from 164 patients. All cases were accompanied by a clinical question of gluten sensitivity and originally interpreted to show normal numbers of duodenal IELs. We reviewed the H&E stained sections from each case and counted IELs and enterocytes in each of 5 villous tips. Cases were immunohistochemically stained with CD3 and CD8, and the number of positive cells in a similar distribution was noted. A mean of ≥12 IELs (or immunopositive cells)/20 enterocytes was considered increased, as described by others. A linear mixed-effects statistical model identified differences in the means determined based on review of H&E stained sections versus immunostained slides.
Results: The study included 52 men and 112 women (mean age: 49 years). Biopsies were interpreted to show no diagnostic abnormality (n=159) or features of peptic injury (n=33). Review of H&E stained slides revealed a mean of 2.09 +/- 0.07 (range: 0.4–5.8) IELs/20 enterocytes, compared to 3.2 +/- 0.1 (range: 0.2–9.6) CD3+ and 2.07 +/- 0.09 (range: 0–6.6) CD8+ cells (p<0.001 and p=1, respectively). This difference was not clinically relevant since none of the cases showed increased IELs. Five patients had elevated TTG antibodies, but mean numbers of IELs were within normal limits by H&E, CD3, and CD8 stains (2.7, 4.2, and 2.6, respectively).
Conclusions: Increased IELs are accurately assessed based on careful evaluation of H&E stained sections and immunostains for T cell markers do not unveil features of gluten sensitivity when routine sections are essentially normal. Thus, efforts should be directed at educating pathologists to recognize the spectrum of histologic features of celiac disease, rather than increasing health care costs with unnecessary stains.
Monday, March 19, 2012 1:00 PM
Poster Session II # 107, Monday Afternoon