A terrific celiac serology study (Gidrewicz D, et al.Am J Gastroenterol 2015; -advanced online publication doi: 10.1038/ajg.2015.87) helps answer questions about the utility of serology in making diagnostic decisions. (Thanks to corresponding author J Decker Butzner for sharing reference.)
Using consecutive samples in a laboratory database with 17,505 patients, the authors retrospectively examined the performance of the tissue transglutaminase (TTG), endomysial antibody (EMA) tests and the ESPGHAN celiac guidelines for nonbiopsy diagnosis of celiac disease. Among this large cohort, 775 with positive TTG and 574 with a negative TTG were biopsied.
- If the TTG >10-fold the upper limit of normal (ULN) along with a positive EMA and symptoms were present, 98.2% had biopsies consistent with celiac disease.
- If the TTG was 3-10-fold-ULN along with positive EMA, then 75.7% had biopsies consistent with celiac disease. The histology of celiac disease (CD) was present in only 40% of the TTG 3-10-fold ULN if EMA was negative.
- If the TTG was 1-3-fold-ULN along with positive EMA, 52.2% had CD-positive histology, whereas CD-positive histology was evident in only 13.3% of TTG 1-3-ULN if EMA was negative.
- IgA deficiency is common in CD (“1in 60 vs 1 in 700” in general population)
Implications & Take-home points:
- The researchers note that the positive predictive value of TTG drops when the prevalence of CD in the testing population is lower. Thus, in their population and most clinical practice where the prevalence is below 35-40%, the PPV of the TTG-based tests drops to <80%.
- While TTG-based tests have high specificity, there are multiple medical conditions that can cause a false positive (additional reference below), including autoimmune diseases like diabetes mellitus, inflammatory bowel diseases as well as infections, and liver disorders.
- Asymptomatic patients with low TTG titers and negative EMA may benefit from following celiac serology rather than proceeding immediately to intestinal biopsy.
- Using ESPGHAN nonbiopsy criteria (symptomatic child, TTG ≥10 ULN, positive EMA, positive HLA typing), there were four patients whose initial biopsies were not consistent with CD. Thus these criteria identified 98.2% accurately who did not need an intestinal biopsy. One of these four developed CD subsequently. To achieve 100% PPV for non biopsy, the authors note that one would need an EMA titer ≥1:80.
- Study limitations: retrospective study, lack of standardization between TTG assays
Bottomline: EMA improves the PPV of TTG testing, especially when low titer elevations are noted. TTG alone is a highly sensitive test “with a 99.4% negative predictive value” in this study.
Related study: “Serum Anti-Tissue Transglutaminase Antibodies Detected during Febrile Illness May Not be Produced by the Intestinal Mucosa” J Pediatr 2015; 166: 761-3. This case report describes two children with abnormal TTG (one more than 20-fold ULN)) both were EMA-negative. No mucosal anti-TTG was identified using two immunoassays.
Related blog posts:
- Nuance in Celiac Serology Interpretation | gutsandgrowth
- Is a biopsy necessary in Celiac disease? | gutsandgrowth
- False-positive serology for Celiac disease | gutsandgrowth
- How Accurate is Serology at Predicting Mucosal Healing in …
- Closer followup for Celiac disease & pediatric guidelines …
- Expert review: Celiac disease | gutsandgrowth