Interlaboratory reproducibility of gastrin measurements by radioimmunoassay.

A collaborative study was conducted to determine interlaboratory variations in the measurement of gastrin by radioimmunoassay (RIA). A kit containing reagents for a communal RIA method, as well as test samples containing different amounts of human synthetic gastrin (SHG I) and test samples of serum from a fasting normal patient admixed with varying volumes of serum from a patient with the Zollinger-Ellison syndrome, was sent to each collaborating investigator. Each investigator was requested to measure the gastrin concentrations in test samples by means of the communal RIA method, as well as the RIA method utilized in his or her laboratory. Use of the communal RIA method resulted in no significant interlaboratory variation in the determination of gastrin concentration in SHG I test samples and a maximum two- to threefold variation in determination of gastrin serum test samples. In contrast, when individual RIA methods were used to measure gastrin concentrations, the amount of SHG I in test samples reported by one laboratory was significantly different than that reported by the other three. In addition, use of different RIA methods resulted in a 2- to 19-fold interlaboratory variation in serum gastrin measurements and the maximum variation was found when the gastrin in normal fasting serum was measured. Increased variations in results, when different RIA methods were used, may be due to differences in preparation and purification of labeled antigen, to differences in separation of bound from free hormone, and to differences in the antibodies themselves.