Non-Biopsy Approach to Transform Adult Celiac Disease Diagnoses

In contrast, European pediatric guidelines have allowed a non-biopsy diagnosis for children with elevated tissue transglutaminase immunoglobulin A (TTG-IgA) antibody levels since 2012, a strategy reaffirmed in their 2020 update. Now, data from a large multicenter study suggest that a non-biopsy approach for the diagnosis of celiac disease may be applicable in a select group of U.S. adults with highly elevated TTG-IgA levels.

The study, conducted by researchers at Cleveland Clinic (Cleveland, OH, USA), involved over 11,000 adult patients across six U.S. sites and sought to determine whether the same non-invasive diagnostic strategy used in European pediatric guidelines could be applied to adults. This large observational study aimed to assess whether a non-biopsy strategy, utilizing a TTG-IgA threshold greater than 10 times the upper limit of normal (ULN) — the same cutoff used in European pediatric guidelines — could accurately identify newly diagnosed cases of adult celiac disease. The study included adults (≥18 years) who underwent duodenal biopsy during esophagogastroduodenoscopy, with TTG-IgA levels measured within three months before or one month after the procedure. Individuals with prior celiac disease diagnoses, total IgA levels below the reference range, or those already following a gluten-free diet were excluded.

Celiac disease was defined histologically by the presence of villous atrophy. Statistical analyses, including receiver operating characteristic (ROC) analysis, were used to evaluate diagnostic performance. Out of 11,282 adult patients assessed across the six centers, most were female (7,739; 68.7%) and non-Hispanic White (8,925; 81.2%). An abnormal TTG-IgA was observed in 1,884 (16.7%) of participants, and villous atrophy was found in 1,800 (16.0%). Patients with higher TTG-IgA values were more likely to exhibit villous atrophy. The sensitivity for any elevated TTG-IgA was 73%, and specificity was 94%. The positive predictive value (PPV) was 71%, and the negative predictive value (NPV) was 95% for celiac disease diagnosis. The test's accuracy was 91%, with a false positive rate of 6%. The area under the curve (AUC) of the ROC curve was 0.88 (95% CI, 0.87-0.89).

Among the participants, 155 patients (8.2%) had TTG-IgA levels greater than 10 times the upper limit of normal. At this threshold, the TTG-IgA test performed excellently, with a specificity of 99.9% and a PPV of 95.5%. A cutoff of TTG-IgA >5x ULN also yielded strong results, with specificity exceeding 99% and a PPV greater than 93%. Notably, only seven of the 155 patients with TTG-IgA >10x ULN did not show villous atrophy on biopsy. These patients were more likely to have a history of other autoimmune conditions (p=0.024). While this large, multicenter study provides valuable insights into non-biopsy diagnosis of celiac disease, researchers emphasized the need for further investigation. High TTG-IgA levels above 10 times the ULN were highly specific for celiac disease, though this threshold applied to fewer than 10% of adults with biopsy-confirmed disease and about 1% of all patients undergoing evaluation at tertiary care centers.

The researchers underscored that biopsy would remain necessary for most adult patients, especially those who are seronegative or IgA deficient. A key next step will be conducting a prospective cohort study to further assess the safety, accuracy, and real-world application of high TTG-IgA thresholds in adult populations. The team also highlighted the importance of including a diverse range of patients to ensure any diagnostic strategy is equitable and effective. This study adds to the ongoing conversation on non-invasive diagnostic methods and highlights a potential pathway toward more personalized care in the future.

“We’re still in a phase where we’re looking to confirm these findings,” said Claire Jansson-Knodell, MD, who led the research. “That’s why prospective validation is so important — this is exciting work, but it needs to be solidly backed before it can shape clinical practice.”