Impact of Assay Epitope Specificity in Gastrinoma

The antral hormone gastrin regulates gastric acid secretion and growth of the gastric mucosa (1 ). As a result of the elaborate cellular maturation process of progastrin, the antral G cells release a mixture of different acidstimulatory gastrins and other precursor fragments to the circulation (2 ). This mixture comprises gastrin-71, -52, -34, -17, -14, and -6, all of which are carboxyamidated and circulate in an O-sulfated and nonsulfated form. The dominant gastrin forms in serum from healthy humans are amidated gastrin-34 and -17 [for a review, see Ref. (3 )]. In cases of increased gastrin synthesis, however, as seen in patients with hypoand achlorhydric gastritis and gastrin-producing tumors, i.e., gastrinomas, the serum or plasma pattern molecular pattern changes to larger molecular forms and more incompletely processed precursors (4–6). Hence, only little gastrin-17 may be present in serum from hypergastrinemic patients, although gastrin-17 is the predominant tissue form in healthy antral mucosa. In clinical chemistry, the only well-indicated routine use of gastrin assays is in the diagnosis of gastrinomas, i.e., the Zollinger–Ellison syndrome (7, 8). Gastrin assays for diagnostic purposes (usually RIA or ELISA) should preferably measure all forms of gastrin. Although assays specific for only one molecular gastrin form may be of interest in biochemical research, such assays pose a significant diagnostic problem. In the following, we illustrate this problem with a newly launched commercial gastrin17-specific ELISA. Serum was obtained from six healthy individuals (three females and three males; median age, 26 years; range, 22–28 years) before and during food intake. Individuals were served a protein-rich meal, and blood samples were collected from the cubital vein before and at set time intervals after the meal. In addition, serum was obtained from seven patients with gastrinomas who had been admitted to the local department of gastrointestinal surgery (five females and two males; median age, 52 years; range, 37–57 years). The diagnosis was ascertained by increased serum concentrations of amidated gastrin, increased basal as well as pentagastrin-stimulated gastric acid secretion, gastroduodenoscopy, selective blood sampling from the portal vein (for gastrin measurements), abdominal computed tomography scanning, and somatostatin analog (octreotide) scintigraphy. All patients later underwent surgery, with the tumor located and histologically verified in six patients. The remaining patient had the tumor identified at autopsy. All healthy individuals and patients gave informed consent for participation in the study, and the local ethics committee approved the study. For measurement of amidated gastrin concentrations in serum, we used a RIA utilizing an antiserum raised against the C-terminal tetrapeptide sequence common to all bioactive gastrins (9 ). This antiserum detects nonsulfated and sulfated gastrin forms with an equimolar potency. N-Terminal gastrin-17 measurements were performed using an antiserum raised against the cyclized NH2 terminus of gastrin-17 (10 ). This assay detects the N-terminal epitope on gastrin-17, where the N-terminal residue is pyroglutamic acid. This assay detects both gastrin-17 and its C-terminal extended precursors. Measurements with a gastrin-17-specific ELISA were performed as suggested by the manufacturer (Biohit Plc, Helsinki, Finland). Briefly, diluted serum was incubated in microplates with a gastrin-17-specific antibody adsorbed to the wells. A second gastrin-17-specific antibody was then added, followed by biotinylated IgG, which bound to the second antibody. Avidin-labeled horseradish peroxidase was used for signal production, and after the enzymatic reaction was terminated, the intensity of color was determined by measuring the absorbance at 450 nm. The between-assay variation (at a gastrin-17 concentration of 30.2 pmol/L) was 7.9%, and the within-assay imprecision (gastrin-17 concentration of 19.6 pmol/L) was 4.7%. The median amidated gastrin concentration in normal serum was 10.5 pmol/L (range, 9.0–18.0 pmol/L), whereas the gastrin-17-specific ELISA measured significantly lower concentrations (median, 4.0 pmol/L; range, 0.0–8.0 pmol/L; n 6; P 0.005). After food intake, the amidated gastrin concentration increased to maximum concentrations 20 min after food intake with a threefold increase in gastrin concentrations (median, 30.0 pmol/L; range, 17.0–79.0 pmol/L). The gastrin-17-specific ELISA detected almost identical concentrations after stimulation (median, 30.0 pmol/L; range, 7.0–96.0 pmol/L). We then examined sera from gastrinoma patients. On average, the gastrin-17-specific ELISA detected only 10% of the amidated gastrin concentrations (Table 1). Consequently, in five of the seven gastrinoma patients, the gastrin-17specific ELISA measured concentrations below the median postprandial concentration. The concentrations obtained with our N-terminal gastrin-17 immunoassay and the gastrin-17-specific ELISA were highly associated, whereas we found no association between amidated gas-