Measurement of faecal alpha 1-antitrypsin: methodologies and clinical application.

Presence of the serum protein, alpha 1-antitrypsin, in the faeces is a potentially useful marker for gastrointestinal disease and/or blood loss. In an effort to simplify faecal sampling procedures, we evaluated the performance of measuring alpha 1-antitrypsin in eluants of thin smears of faeces made on filter paper, relative to conventional measurement in aqueous extracts of stool. Faecal specimens and smears were collected from healthy subjects (n = 22) and from patients with gastrointestinal bleeding (n = 12) or inflammation (n = 22), colorectal neoplasia (n = 15), or miscellaneous diseases with a low risk of excessive faecal protein loss (n = 30). alpha 1-Antitrypsin was measured by ELISA and haem porphyrins (as a measure of blood loss) by HemoQuant. Results from smears were highly correlated (r = 0.81; P < 0.001) with faecal alpha 1-antitrypsin. The smear method detected an elevated faecal alpha 1-antitrypsin with 93% specificity and 75% sensitivity. Sensitivity was high ( > 88%) where levels were markedly elevated in inflammatory and bleeding groups and low ( < 62%) where abnormal levels were mildly elevated (neoplasia and miscellaneous groups). Elevated alpha 1-antitrypsin detection by either method positively predicted > 90% of patients with gastrointestinal inflammatory disease when levels were elevated 6-fold and 1.5-fold or more, respectively. In 15 patients with colorectal neoplasia, faecal alpha 1-antitrypsin was elevated in 10 patients, haem porphyrins in nine patients and either in 12; however, smear eluant levels were elevated in only six patients. Blood loss was probably a major contributor to elevated faecal alpha 1-antitrypsin in some patients but not in the inflammatory group as a whole. The sampling and aesthetic advantages of the smear eluant method are offset by reduced sensitivity, precluding its use as a screening test for colorectal neoplasia. However, its performance in predicting inflammatory disease is equivalent to that of conventional measurement and warrants a prospective evaluation as an early investigative test. Concurrent evaluation of blood loss may improve its interpretation.