|| Checking for direct PDF access through Ovid
To the Editor: We thank Dr Tortora and colleagues for the positive and interesting comments (1) in response to our recently published manuscript on gastrointestinal symptoms and pathology in common variable immunodeficiency (CVID) (2). One of our assertions was that CVID patients with increased intraepithelial lymphocytes—celiac-like disease—were different from ‘true’ celiac disease (CD) patients in terms of both gene expression array and human leukocyte antigen (HLA) profile.Diagnosing CD in CVID patients is challenging due to the unreliable autoantibody testing in patients with B-cell deficiency, incoherent HLA profile (3, 4), and inconsistent response to a gluten-free diet (GFD) (3). Tortora et al. suggest two additional approaches in evaluating CD in CVID; (i) Assessing T-cell immune response to in vitro gliadin challenge of duodenal biopsies, and (ii) investigating gluten-dependent TG2-specific immunoglobulin A (IgA) deposits in the small bowel mucosa. Although interesting, both approaches involve an assumption of normal immune responses. However, studies have shown abnormal T-cell response in CVID patients, including a reduced number of total, naive, and regulatory CD4+ T cells, as well as functional impairment such as reduced stimulated proliferation and altered production of cytokines (5, 6). As part of their hypogammaglobulinemia, most CVID patients do not produce autoantibodies in plasma, which is actually included as a diagnostic criterion by some (7). Thus, neither a reduced T-cell response to gliadin challenge nor the lack of TG2-specific IgA deposits in the small bowel mucosa can totally rule out CD in CVID because these findings may be due to the immunodeficiency per se and not related to reduced response to a specific insult like gluten. However, a positive finding of either T-cell response (possibly by direct demonstration of pathogenic T cells (8)) and/or TG2-specific IgA deposits may support CD in these patients, and the results of such studies could be of interest.We agree that the only criterion to confirm CD diagnosis in CVID patients is the histological response to a GFD. However, it has also been suggested that HLA genotyping can be useful in selecting which CVID patients with celiac-like histopathology should be tested with a GFD, but, to this end, the supporting data available are scarce and conflicting (3, 4).The difficulties in diagnosing CD in CVID may not only be due to the immunodeficiency making diagnostic tools unreliable, but can reflect that celiac-like findings in CVID, in most patients, involve a different disease mechanism and are not triggered by gluten or infection. Instead, the celiac-like findings in CVID can be a result of an immune dysregulation affecting the duodenum and, in most cases, also other parts of the gastrointestinal tractus.