Abstract Background Celiac disease (CD) screening guidelines recommend testing symptomatic patients older than three years with IgA anti-tissue transglutaminase (IgA anti-TTG) and total IgA. Testing for anti-deamidated gliadin peptide (anti-DGP) IgG, along with anti-tTG IgG, is recommended for individuals with low IgA or IgA deficiency. According to the recommended reflexive testing algorithm, detectable total IgA concentration and negative IgA anti-TTG results rule out CD. In our laboratory, we utilize a “nonreflexive” testing algorithm that reports all available Celiac serology results including anti-TTG IgA, anti-DGP IgG and IgA, and total IgA, concurrently. While this approach appears to enhance the sensitivity of detecting CD in symptomatic patients, concerns about test overutilization arise. However, a reflexive testing algorithm may potentially miss some CD patients who have normal anti-TTG IgA results. As the recommendation to begin initial CD screening with IgA anti-TTG and total IgA gains momentum, it is important to evaluate the performance of Celiac serologies against diagnostic gold standards to ensure that clinically relevant cases are not missed. This is particularly crucial, as Celiac disease immunoassays are not standardized. This study aims to compare the performance of the Thermo Fisher Phadia Elia immunoassay CD biomarkers to biopsy findings in our patient population. Methods Phadia Elia immunoassay data were collected retrospectively from 2019-2023 from our electronic medical record. A total of 770 eligible patient results were identified. Of the 770 cases only patients with recent biopsy results with pathologist interpretation and negative anti-tTG IgG were included in the study (N = 240). These patients would have been considered negative according to the reflexive algorithm. Continuous data are presented as mean ± SD. Results The demographics of the study population were majority female (60.3%), plurality Hispanic (41.4%) and had a mean age of 40.16 ± 23.63 (total range 3-91). Of the 240 patients, 36 (15%) cases were diagnosed with positive biopsy results, of which 8 (22.22%) cases had other concomitant autoimmune disorders. The IgA Anti-DGP in the positive biopsy group ranged from 10 U/mL to 100 U/mL (mean 25.11 U/mL ± 23.94) and the IgG Anti-DGP ranged from 10 U/mL to 100 U/mL (mean 25.11 U/mL ± 30.97). There were 204 (85%) cases with negative biopsy results, of which 23 (11.3%) had other concomitant autoimmune disorders. The IgA Anti-DGP of the negative biopsy group ranged from 10 U/mL to 100 U/mL (mean 21.65 U/mL ± 17.43), and the IgG Anti-DGP ranged from 10 U/mL to 46 U/mL (mean 20.99 U/mL ± 9.9). Conclusion In conclusion, the current reflexive screening guidelines for CD lack sufficient sensitivity for our patient population, as 36 clinically actionable cases would have been missed if the recommended algorithm had been adopted.
Gill et al. (Wed,) studied this question.