Definitive diagnosis of gluten-sensitive enteropathy. Use of an in vitro organ culture model.

The flat mucosal lesion of the small intestine is not pathognomonic of gluten-sensitive enteropathy (GSE). Frequently, the definitive diagnosis of this condition can only be established after three intestinal biopsies are performed: an initial one to show a flat mucosal lesion, one after a gluten-free diet to show morphological recovery, and one after a gluten challenge to show morphological deterioration. We used an organ culture model of GSE to determine the usefulness of this technique in establishing a diagnosis of GSE on the basis of the initial biopsy. Seventy-five patients with diarrhea, and/or malabsorption were evaluated prospectively; 40 had a flat mucosal lesion of variable degree; of these 26 were ultimately determined to have gluten-sensitive enteropathy by the above criteria. A rise in alkaline phosphatase activity of intestinal tissue from 22 of these 26 patients was inhibited when the tissue was cultured in gluten-containing medium as compared to enzyme activities of cultures in a gluten-free medium (108 +/- 69 versus 206 +/- 96, mean +/- SD, P less than 0.001). Mean enzyme values in the similarly cultured intestinal tissue from 13 of 14 patients ultimately shown not to have GSE were not affected by gluten (224 +/- 94 versus 201 +/- 109, P greater than 0.4). Examination of the data by stepwise discriminant analysis provided a function which correctly classified 35 of the 40 patients (88%). The false-positive and false-negative rate for establishing the diagnosis of GSE was 7% (1 of 14) and 15% (4 of 26), respectively. All patients with normal biopsies were classified correctly. The model can be used to establish prospectively the definitive diagnosis of GSE, obviate the need for additional diagnostic biopsies, and allow for the prompt pursuit of alternative diagnoses when gluten sensitivity is not shown.