Association of serum antibodies against p53 protein with poor survival in patients with Zollinger-Ellison syndrome.

BACKGROUND & AIMS: Long-term survival of patients with Zollinger-Ellison syndrome (ZES) is largely determined by the presence or absence of liver metastases. However, because of the lack of precision of these criteria, development of further indicators is still required. Recent evidence showing that autoantibodies directed against the p53 protein could predict poor survival for some types of cancers prompted us to investigate the presence of such antibodies in sera from patients with ZES and their potential value as survival indicator. METHODS: Anti-p53 antibodies were detected in the sera of 44 consecutive patients with ZES using both an enzyme-linked immunosorbent assay (ELISA) and Western blotting. The mean follow-up of these patients was 92 months. RESULTS: Anti-p53 antibodies were detected in 7 of the patients with ZES (16%) by both ELISA and Western blotting. Univariate and multivariate analyses showed that the presence of anti-p53 antibodies (P = 0.0009 and P = 0.017, respectively) and liver metastases (P = 0.0009 and P = 0.012, respectively) was independently associated with shorter survival. CONCLUSIONS: These results suggest that anti-p53 antibodies are an indicator of survival and could be used in combination with staging to determine which patients with ZES have poor prognoses and therefore require reinforced therapy.

The effect of IL-4 on human nasal mucosal responses.

Interleukin (IL)-4 causes the dose limiting sensation of nasal congestion when administered systematically at doses of 3 micrograms/kg or higher thrice daily to humans. This side effect was observed in a group of patients treated as part of an immunotherapy protocol for cancer management. To determine the source of this congestion, nasal secretions were collected prospectively in a group of patients at baseline and after provocation with normal saline, methacholine (which stimulates glandular secretion), and histamine (which causes increased vascular permeability). Nasal lavages obtained at baseline and after provocation were analyzed for the presence of these glandular and vascular proteins and inflammatory mediators. Washings and provocations were performed before IL-4 administration, after 24 hours of IL-4 treatment, and after 3 days of treatment, at a time when nasal congestion was maximal. Compared with histamine challenge before IL-4 treatment, the secretion of the plasma proteins albumin and IgG were significantly decreased after 3 days of IL-4 treatment. IL-4 treatment had no apparent effect on methacholine-induced responses. Thus systemically administered IL-4 causes the subjective sensation of nasal congestion, increased histamine in nasal lavages, and the development of vascular unresponsiveness to histamine, without affecting parasympathetic responses to histamine. The relationships among increases in nasal lavage histamine, vascular unresponsiveness to histamine, and the sensation of nasal congestion are unclear.

Could chronic peptic ulcers be localised areas of acid susceptibility generated by autoimmunity?

It is argued that all chronic gastroduodenal peptic ulcers result from localised increase in mucosal susceptibility to acid attack at the interface between a segment of gastroduodenitis and gastric fundus or duodenal mucosa. The site is predetermined by the background mucosal pattern. Changes can occur in the differentiated gastroduodenal mucosa that closely resemble cell population transformations described in embryology and regeneration biology. A second pathological process, gastroduodenitis, may develop that does not of itself predispose to ulceration, but the combination of factors can produce a zone of increased acid susceptibility. These complex changes could be generated by immunologically activated gastroduodenitis. Destructive or stimulatory immune reactions, analogous to those seen in the thyroid gland, could affect the gastrin-secreting G cells and other paracrine cells. The resulting tropic and inflammatory reactions would provide the background for peptic ulceration.

Specificity of commercially available antibodies used for gastrin measurement.

We examined the specificity of commercially available antibodies used in measurement of serum gastrin. Antibodies were obtained from five commercial laboratories, and antibody immunoreactivity with gastrin and cross-reactivity with cholecystokinin (CCK) were determined. All antibodies were equally immunoreactive with gastrin, and cross-reactivity of three antibodies with CCK was minimal (less than 5%). In contrast, substantial cross-reactivity with CCK was found with two antibodies. To determine the clinical significance of cross-reactivity with CCK, secretin injection tests were performed in 24 individuals: seven in normal health, four with Zollinger-Ellison syndrome, three with antral gastrin cell hyperfunction, six with ordinary duodenal ulcer disease, and four with atrophic gastritis. Serum gastrin levels were measured with all five gastrin antibodies. The response to secretin was negative in all normal subjects and in those with duodenal ulcer and antral gastrin cell hyperfunction. The response to secretin was positive in all four patients with gastrinoma with use of the five antisera. All four patients with atrophic gastritis had normal responses to secretin when antibodies with minimal CCK cross-reactivity were used; however, two of four had false positive secretin test results when serum gastrin levels were measured with the two antibodies with a high degree of cross-reactivity with CCK. These studies indicate that significant cross-reactivity of gastrin antibodies with CCK can result in false positive secretin injection test results and can lead potentially to the erroneous diagnosis of Zollinger-Ellison syndrome.

Localisation of C-terminal gastrin immunoreactivity in gastrinoma cells. An immunoelectron microscopy study on conventionally processed tissue.

Localisation of C-terminal gastrin immunoreactivity has been studied, using the immunogold staining procedure, on ultrathin sections of 6 human gastrinomas conventionally processed for electron microscopy. The specific labelling, whose density depended on the mean diameter of the gold marker, was restricted to endocrine secretory granules. However, in poorly differentiated cells from malignant tumours, a number of granules remained unreactive. The labelling pattern depended also on the functional state of each cell. The immunoreactive granules showed various morphological features. A moderate number of gold particles was demonstrated over the floccular content of the infrequent diagnostic G-type granules. Non-diagnostic round granules of varying size and electron density were prevalent in most cells; their usually strong immunostaining allowed immediate recognition of cell specificity. Dense granules which were large in size and angular in shape and present in one case, were also intensely labelled. In the same tumour, unequal labelling occurred over polymorphous, often elongated granules, of varying size. Granules of different types, including intermediate forms, could be found in the same cell, indicating a spectrum of granule maturation towards well-defined types of the fetal or adult normal tissues. The present methodology would help to identify gastrin-producing cells in prospective or retrospective electron microscopy studies of multi-hormonal endocrine tumours.

Pancreatic endocrine tumors.

One hundred twenty-five pancreatic endocrine tumors were analyzed by immunocytochemistry using various antisera. Twenty-three of 27 insulinomas, 10 of 10 PP-omas (PP: pancreatic polypeptide) and 15 of 30 "nonsecreting" tumors were benign, whereas 8 of 13 glucagonomas, 16 of 24 gastrinomas, and 16 of 21 VIP-omas (VIP: vasoactive intestinal polypeptide) were malignant. As a rule, the hormone secreted by the tumor and causing clinical symptoms could be localized by immunocytochemistry. Fifty of 95 active tumors were found to contain cells immunoreactive to peptide(s) not causing clinical symptoms, and 54 of 30 "nonsecreting" tumors were shown to be multicellular. By electron microscopy more than one cell type could be identified in 12 tumors. Histologically, the growth pattern of the tumors was very variable and distribution of immunoreactive cells was distinctly patchy. Radioimmunoassay on extracts of 20 of 27 tumors confirmed the presence of peptides visualized by immunocytochemistry. In 17 of 22 specimens, groups of endocrine cells in close contact with ductules were found in the pancreatic parenchyma distant from the tumor. Pancreatic endocrine tumors probably arise from the pancreatic ductular epithelium. They are often multicellular, producing and sometimes secreting more than one hormone or hormone-like substance. They represent highly complex biologic systems in which the interrelationship of various gastrointestinal-pancreatic hormones can be studied.

Parietal cell autoantibodies and hypergastrinemia in achlorhydria and the Zollinger-Ellison syndrome.

Parietal cell autoantibody (PCA), basal gastrin, and calcium-stimulated gastrin were measured in twenty patients with achlorhydria, in eight patients with the Zollinger-Ellison syndrome, and in fifty control subjects. In twelve patients with achlorhydria with a spared antrum, PCA was positive and basal gastrin was elevated. In contrast, eight achlorhydric patients with antral gastritis had negative PCA and significantly lower basal gastrin levels. Patients with the Zollinger-Ellison syndrome did not demonstrate positive PCA despite elevated levels of basal gastrin, nor was PCA present in normal controls. This study suggests that certain achlorhydric states are caused by an autoimmune response, particularly if antral function is spared.

Heptadecapeptide gastrin: measurement in blood by specific radioimmunoassay.

The characteristics are described of an antibody (designated L6) which has virtually absolute specificity for heptadecapeptide gastrin. This antibody binds G17, but does not bind peptide fragments or molecular forms of gastrin comprising G17 with either amino acid deletions, or additions, at the carboxyl- and amino-terminals. In serum from patients with Zollinger-Ellison syndrome the only form of gastrin revealed by L6 was compatible with G17, and there was good agreement between estimated G17 concentrations in serum analyzed by gel filtration and by direct radioimmunoassay using L6. Using L6 in conjunction with antibodies specific for carboxyl- and amino-terminals of G17 it has been possible to measure concentrations of different forms of gastrin in serum of normal subjects after a meal in greater detail than previously possible. After a light meal consisting of eggs, toast, and Oxo, serum concentrations of G17 measured by L6 increased to a peak 20 min after feeding (delta gastrin, 19 pmoles per liter; n = 17). In contrast, concentration of G34 peaked at 50 min (delta gastrin, 27 pmoles per liter). Small amounts of amino-terminal fragments of G17 were present throughout the digestive period. Applying the known ratio of biological potencies of G34 and G17 for stimulation of acid secretion in man, it is estimated that G17 accounts for about 75% of the biological activity in blood after a meal, even though G34 is present in higher molar concentrations.

Therapy with gastrin antibody in the Zollinger-Ellison syndrome.

The therapeutic effectiveness of parenterally administered rabbit antigastrin antibody was evaluated in a patient with the Zollinger-Ellison syndrome who had a fasting serum gastrin level of 3020 pg/ml and a basal gastric acid secretion of 48.9 mEq/hr. Control globulin reduced gastric secretion to 32 mEq/hr. Gastrin antibody reduced it futher to 8.7 mEq/hr. Betazole hydrochloride which was given 75 min after administration of gastrin antibody stimulated acid secretion to 57.2 mEq/hr. One day later basal acid secretion was uninhibited although some antibody activity was present in the patient's serum. The results suggested that gastrin antibody acutely inhibited basal but not betazole-stimulated secretion.

Amino terminal gastrin fragment in serum of Zollinger-Ellison syndrome patients.

In addition to the previously described molecular forms of gastrin, a new component has been found in high concentrations in the serum of 6 patients with Zollinger-Ellison syndrome. Serum was fractionated by gel filtration, and ion exchange chromatography, and the new component was identified in eluates by radioimmunoassay using an antibody with specificity for the N-terminal portion of heptadecapeptide gastrin. The precise chemical nature of the new component is not known, but its chromatographic behavior and its reactivity to various antibodies is indistinguishable from that of the natural or synthetic N-terminal 1 to 13 fragment of G-17. The new component is present in gastrinoma tumor tissue. Its concentration in serum of gastrinoma patients increases markedly when secretin is injected.

Isolation of two minigastrins from Zollinger-Ellison tumour tissue.

A pair of gastrin tridecapeptides (;minigastrins') have been isolated from Zollinger-Ellison tumour tissue; they correspond to the fragment 5-17 of the heptadecapeptides isolated from the same source. In one (type II) the tyrosine residue present is sulphated, in the other (type I) it is not. The proportion of types I and II is approximately 2:1 similar to that for the ;big' gastrins and the heptadecapeptides isolated from the same source and from human antral mucosa. Both minigastrins are potent stimulants of gastric acid secretion. Immunological evidence exists to indicate that both are present as circulating forms of the hormone gastrin in patients with the Zollinger-Ellison syndrome and in fed normal subjects.

Immunoreactive gastrin components in human serum.

The apparent molecular size and charge of immunoreactive gastrin components were studied in sera from patients with pernicious anaemia or gastrinomas (the Zollinger-Ellison syndrome) by Sephadex gel filtration and aminoethylcellulose chromatography. The following serum components were distinguished: (1) a monophasic component I similar in size to proinsulin which was converted into ;little' gastrin I by trypsin digestion; (2) a biphasic component II, corresponding to ;big' gastrins I and II (Gregory and Tracy); (3) a biphasic component III corresponding to ;little' gastrins I and II (Gregory and Tracy); and (4) a biphasic component IV, corresponding to ;minigastrins' I and II (Gregory and Tracy). ;Big, big' gastrin, a plasma component found in the void volume of the Sephadex G-50 column by Yalow and Berson (1972) was undetectable in the sera investigated. A component in gastrinoma and antral mucosa extracts corresponding in size to ;big big' gastrin was detectable by the assay; the ;big big' gastrin fraction from gastrinoma tissue was heterogenous, with components of apparent MW 30 000-100 000. It is concluded that serum gastrin circulates in the form of at least four components, of which the three smaller ones are in pairs.

Gel filtration studies on immunoreactive gastrin in serum from Zollinger-Ellison patients.

Sera from 15 patients with the Zollinger-Ellison syndrome were subjected to gel filtration on Sephadex G-50 superfine columns (10 x 2000 mm). The concentration of gastrin in the effluent was determined by a sensitive radioimmunoassay. Immunoreactive gastrin was eluted in four components in 14 sera. (1) Component I, eluted in the same position as proinsulin, constituted 9.7 +/- 1.2 (mean +/- SEM)% of the total immunoreactivity. (2) Component II (;big gastrin') eluted between proinsulin and insulin constituted 57.8 +/- 4.1% (mean +/- SEM) of immunoreactive gastrin. In three sera with the highest concentration of gastrin, component II appeared biphasic. (3) Component III (;little gastrin') was distributed in two peaks; the first one eluted in the same position as the heptadecapeptide gastrin II made up 17.4 +/- 2.7 (mean +/- SEM)% of the total immunoreactivity; the second one eluted in the same position as gastrin I constituted 9.5 +/- 1.3 (mean +/- SEM)%. (4) Component IV (;minigastrin') was eluted immediately before the salt peak and constituted 5.6 +/- 1.4 (mean +/- SEM)%. In one serum only components I and II were present. After incubation with trypsin all immunoreactivity in components I and II was converted to heptadecapeptide-like gastrins.The findings suggest that immunoreactive gastrin in serum from Zollinger-Ellison patients is circulating in at least four components of different molecular size.