Thirty human insulinomas have been investigated histologically and their immunoreaetive insulin (IRI) content estimated. In most eases immunohistological and ultrastruetural studies were also performed and the percentage of proinsulin-like components (PLC) in the tumour determined. Except for 1 ease the II~I concentration in the tumours was lower (0.01-89.0 U/g) than in the islet tissue. Histologically, immunohistologically and ultrastrueturally a variable number of tumour cells contained few and often no beta-granules, indicating a decreased storage capacity for insulin. This defective storage capacity seems to be the major functional abnormality of insulinoma cells. Ultrastrueturally four types of insulinoma can be distinguished. The ultrastructural diagnosis of an insulinoma can only be made in t?~e I (typicai beta-granules, 13 cases) and type II (typieM and at~2pical granules, 7 eases) but, not in type III (atypical granules only, 4 eases) and type IV (virtually agranular, 4 eases). The type IV tumours had the lowest IRI concentration and did not respond to diazoxide treat-merit. The IRI concentration of the uninvolved pancreas of 19 patients was 2.0~0.2 U/g and in the range of nondiabetic adults.-The percentage PLC in 19 insulinomas was higher (5.3-22%) than in the pancreas of human adults with and without insulinoma (1.7-4.8%). The percentage of PLC in the serum of patients with insulinoma was always higher than in their tumours (33-61%). It is suggested that the higher PLC levels found in the tumour and serum of insulinoma patients are the consequence of the reduced storage capacity of the tumour cells resulting in a rapid passage through the granular route or even a non-granular release of newly synthesized insulin.

SUMMARY The mean antral immunoreactive gastrin (IRG) concentration of 38 duodenal ulcer (DU) patients was significantly higher (35.9 ± 5 2 jg/g) than that of 21 controls (15.9 + 2-6 jg/g). Also the mean IRG concentration in the proximal duodenal mucosa of 15 DU patients (3.2 ± 0O8 jg/g) was higher (but not significantly) than that of 10 controls (1-8 0S5 jg/g). The number of G-cells in the antral mucosa of 58 DU patients and in the duodenal mucosa of 29 DU patients was not larger than that of controls. The distribution of immunoreactivity in gastrin components has been investigated in the antral and duodenal mucosa of six DU patients and six controls. In the antral mucosa the mean percentage of G-17 was 93 3 % in DU patients and 920 0% in controls. G-34 amounted to 4-0°% in DU patients and to 5-0% in controls. The G-34 percentage in the duodenal mucosa was higher (however not significantly) in the DU patients than in the controls (50 1 % versus 35 8 %). Ultrastructurally, the antral G-cells of DU patients had a significantly lower density index of their secretory granules suggesting higher functional activity. It is concluded that the exaggerated serum IRG response of DU patients to different stimuli is not a consequence of an increased G-cell mass.

No abstract

Abstract. In extracts of human gastric biopsies gastrin has been estimated with an immunochemical method. In the same biopsies G‐cells have been localized with an immunohistological method using peroxidase‐labelled antibodies and the endocrine cells investigated electron‐microscopically. Gastrin and G‐cells could be found regularly in the antral mucosa and only in insignificant amounts or not at all in the fundic mucosa of six normal persons. With the same methods gastrin and G‐cells could be demonstrated in the antral mucosa of rats and guinea‐pigs. The gastrin content of the antral mucosa of six patients with pernicious anaemia and achlorhydria with elevated serum gastrin levels was more than 20 times higher than in the controls and the G‐cells were significantly more numerous. Besides hyperplasia of the G‐cells, increased secretory activity was found electron‐microscopically. The gastrin release from the G‐cells seems to take place mainly via intracellular dissolution of the granule content within the membranous sacs. Although the number of other endocrine cells was increased in pernicious anaemia the ultrastructural identity of the G‐cells could be established by comparison with the cells of a Zollinger‐Ellison tumour. This tumour contained gastrin and gave a positive immunohistological reaction for this hormone. Also, the fundic mucosa of patients with pernicious anaemia contained gastrin and G‐cells, but considerably less than the antral mucosa. Hyperplasia of G‐cells was found in six cases of acromegaly, four of which also had a significantly increased gastrin content of the antral mucosa. This finding suggests a trophic function of the hypophysis, especially growth hormone, on the G‐cells. Hyperplasia of the G‐cells in the antral mucosa of three patients with primary hyperparathyroidism and increased gastrin content in two of the three cases also suggest a trophic function of the serum calcium level on the G‐cells.

The recent elucidation of the structure of apo(a), the characteristic apoprotein of the lipoprotein Lp(a), has revealed a high homology to human plasminogen [6, 8]. This similarity provides a tentative link between the plasma lipoproteins and the clotting system and may give further insight into the role of Lp(a) in atherosclerosis [2]. Numerous studies have demonstrated a strong association between high levels of Lp(a) and coronary heart disease [1, 3, 10]. Furthermore, there is an additive effect of LDL and Lp(a) such that the relative risk for angiographically documented coronary artery disease at Lp(a) concentrations of 30 mg/dl and greater increases from twofold to fivefold when LDL are also elevated [1]. Whereas both LDL and Lp(a) levels can be lowered by neomycin and niacin [4], bile-acid resins such as cholestyramine do not affect Lp(a) concentrations [11]. The HMG-CoA reductase inhibitors, have proven to be extremely effective in lowering plasma LDL and apo B levels [5, 9], presumably through inhibition of intracellular cholesterol synthesis with a concomitant increase in the livers LDL receptors [7]. In order to establish whether Lp(a) levels are also After a wash-out phase of 4 weeks in which all previous lipid-lowering therapy, with the exception of diet, was stopped, serum total cholesterol, LDL-cholesterot, apoB and Lp(a) concentrations were determined. Total cholesterol was measured with an enzymatic test kit, (Boehringer, Mannheim), LDL-cholesterol by a specific precipitation method (Quantolip-LDL, Immuno, Heidelberg), apoB by rate nephelometry and Lp(a) with a specific non-competitive ELISA. All individuals were administered a single daily dose of 10 mg Simvastatin for 6 weeks and this was then raised to 20 mg once daily for a further 6 weeks. Lipid and lipoprotein parameters betbre and after therapy are summarised in table 1. Simvastatin caused a significant reduction in mean total cholesterol (32%), LDL-cholesterol (38%) and apoB (26%) concentrations. No consistent effect was observed on serum Lp(a) concentrations and both the mean and median remained unchanged after 12 weeks treatment. These results confirm that Lp(a) levels are not altered by increasing the LDL-receptor activity.