The diagnosis of CD rests on histologic confirmation with duodenal biopsy and a positive response to a gluten-free diet.2 Serologic tests are critical for CD screening and should be considered for first-degree relatives of patients with a CD diagnosis. Typical findings include crypt hyperplasia, intraepithelial lymphocytosis, and villous atrophy. These findings are characteristic of CD, but not specific or diagnostic. Ultimate confirmation is provided by clinical improvement on a gluten-free diet.
Before undergoing endoscopy with biopsy, most patients will have serologic testing as a screen. Immunoglobulin A (IgA) antibodies offer the most sensitive testing, and endomysial serology offers the greatest specificity, with near 100% accuracy.2 Tissue transglutaminase is the auto-enzyme that produces endomysial antibodies, and currently available enzyme-linked immunoassays offer a rapid, less-expensive quantitative alternative to endomysial antibody testing. Both tests have greater than 90% sensitivity, are considered optimal for screening for CD, and correlate with the extent of mucosal injury. Antigliadin antibody testing, once recommended for screening, lacks the sensitivity and specificity for use as a screening test in reproductive-aged patients.
Selective IgA deficiency is 10 times more common in patients with CD, and affected patients may not have abnormal antibody testing.2 For this reason, some investigators have recommended that anti-tissue transglutaminase antibodies be used as the sole screening test and, if not elevated and CD is strongly suspected, total IgA levels should be obtained to exclude the diagnosis of selective IgA deficiency.2,20 IgG antibodies to tissue transglutaminase are often elevated in this setting.
Testing of serum IgA anti-tissue transglutaminase antibodies is recommended initially in those without IgA deficiency, because it is highly sensitive and specific.1 IgA anti-endomysial antibodies are nearly 100% specific but may only be used to confirm diagnosis, because they may be seen in other autoimmune disorders.
Ninety percent to 95% of patients with CD carry the HLA-DQ2 allele and almost all the additional patients have HLA-DQ8.2 Approximately 30% to 40% of the population carries these genes, so testing for them offers poor predictive value. Their absence makes the diagnosis highly unlikely, and they may be occasionally useful as a negative screening test in people who are at risk.