Characterization of the C-terminal region of molecular forms of human milk bile salt-stimulated lipase.

UNLABELLED: Bile salt-stimulated lipase (BSSL) in human milk exists in multiple molecular forms and it has been shown that approximately one-third of lactating mothers secrete two forms. AIM: to determine the structural features of BSSL that may give rise to this heterogeneity. METHODS: Oligosaccharides present in the proline-rich region in the C-terminus of BSSL were investigated using deglycosylating enzymes and lectin affinity probing to determine the origin of the multiple molecular forms. RESULTS: It was found that the variability in the molecular mass of BSSL is due predominantly to glycosylation. The molecular forms contain similar sugar chains; all forms possess the core disaccharide Galbeta1-3GalNAc and beta-D-galactose, fucose linked at alpha1-6 and sialic acid linkage alpha2-3 to galactose. CONCLUSION: The molecular mass difference in the BSSL molecular forms cannot be attributed to the type of carbohydrate moiety in the sugar chains of the N- and O-linked sites suggesting that the differences arise from the extent or quantity of glycosylation. The oligosaccharides in the C-terminal region contain Lewis x and b and, less prominently, Lewis a antigenic structures. Owing to the presence of these blood-group-related antigenic determinants, the C-terminal region of BSSL may have an adhesive function in cell-cell interactions.

Tobacco smoke exposure in children and adolescents with diabetes mellitus.

AIMS: To examine active and passive tobacco smoke exposure in children and adolescents attending a diabetic clinic. METHODS: Salivary cotinine concentrations were measured by gas chromatography and questionnaire data on the smoking habits of patients, families and friends were analysed as well as recording of glycosylated haemoglobin (HbA1c), body mass index (BMI) and social deprivation score. RESULTS: Salivary cotinine concentrations identified 7% of the patients as active smokers and 72% as passive smokers. The mean cotinine concentration in those with no identifiable source of exposure was 0.10 (95% confidence interval 0.05-0.14) ng/ml, 2.81 (2.24-3.38) ng/ml in the passive smoking group and 1003.69 (55.96-151.41) ng/ml in the active smokers. Cotinine concentrations in passive smokers increased with the number of sources of exposure. The mean cotinine concentration was also higher when the mother was the sole source compared to other sources. There was no statistically significant correlation to smoking exposure and HbA1c BMI and deprivation scores. CONCLUSION: Tobacco smoke exposure may pose serious health risks to children and adolescents with diabetes and additional public health measures are required to reduce overall exposure.

Incidence of molecular forms of bile salt-stimulated lipase in preterm and term human milk.

Preterm and term human milk samples obtained at various times after delivery were analyzed for the presence of molecular forms of the human milk enzyme, bile salt-stimulated lipase (BSSL). Thirty-five percent of both the preterm and term milk samples contained two molecular forms of BSSL, of variable molecular mass. The remainder contained only one molecular species of either 115 kD (50%) or 120 kD (15%). The number of molecular forms present was not related to length of lactation, maternal age, gestation, or maternal blood group. The specific activity of BSSL purified from term milk was similar to that purified from preterm milk, and there was no difference in specific activity whether one or two molecular forms were present. This study demonstrates heterogeneity of both molecular mass and molecular forms. We conclude that preterm babies fed their own mother's milk are unlikely to be disadvantaged with respect to fat digestion as BSSL secreted in preterm milk appears to be very similar to that produced in term milk, although we cannot exclude other functional differences.

Lyso-phosphatidylcholine and outcome of preterm babies with respiratory distress syndrome treated with surfactant.

Phosphatidylcholine (PC) is the predominant phospholipid in natural surfactant preparations. A metabolic intermediate, lyso-PC, is potentially injurious to the lungs. In the present study, tracheal aspirates from preterm babies with respiratory distress syndrome treated with surfactant were examined for the presence of lyso-PC to determine if there was any correlation with outcome. Eighteen babies were assigned to receive initially either 100 or 200 mg/kg Curosurf followed by up to three further 100-mg/kg doses if required. Lyso-PC was present in aspirates taken 12-24 h after the last treatment from nine of 11 infants who initially received 200 mg/kg but in only one from seven receiving 100 mg/kg initially, and was dependent on the total dose of phospholipid administered. Three babies in the low-dose group developed bronchopulmonary dysplasia, whereas two in the high-dose group were non-survivors, however we could not correlate the presence of lyso-PC with adverse long-term outcome in this group of preterm infants.

Biosynthesis of neuropeptide Y in porcine tissues and generation of N-terminal fragments in neuroblastoma cell lines.

The biosynthesis of neuropeptide Y (NPY) was investigated to determine the efficiency of synthesis and processing of the precursor. In brain tissues examined, the major product was amidated NPY(1-36). Although this was also the major product in adrenal and heart atrium, a minor portion of the immunoreactivity was identified as unprocessed precursor. NPY degradation was investigated using SK-N-MC and SMS-MSN cells in conjunction with iodinated NPY tracers, labeled in either the tyrosine-1 or the tyrosine-36 position. Similar patterns of degradation were observed with both cell lines, and it would appear that the initial proteolytic attack on NPY(1-36) generates predominately N-terminal fragments.

Stated versus actual lipase activity in pancreatic enzyme supplements: implications for clinical use.

We have carried out an independent investigation of the lipase activity of several pancreatin preparations, including high-lipase preparations of pancreatin that have recently become available. Six preparations from three pharmaceutical companies, Cilag, Duphar, and Merck, were analyzed during the recommended shelf life of the preparation and included a standard and a high-lipase preparation from each manufacturer. All preparations studied showed lipase activity in excess of that which was stated on the packets, and in some cases it was more than twofold. This has important implications for patients and prescribers since changes in apparent enzyme requirements may reflect differences in the potency of batches with time. As it is impossible to evaluate capsule requirements with each new batch prescribed, the presence or relief of symptoms of malabsorption will probably continue to be the way in which most patients monitor their enzyme supplementation on a regular basis. We believe that clinicians need to be aware of the extent of possible intrinsic variation between individual prescriptions before attempting to define possible dose/symptom relationships.

Processing of two homologous precursors, pro-neuropeptide Y and pro-pancreatic polypeptide, in transfected cell lines expressing different precursor convertases.

The processing of two homologous precursors, pro-neuropeptide Y (pro-NPY) and pro-pancreatic poly-peptide (pro-PP), was studied in four neuroendocrine cell lines after transfection: CA-77 medullary thyroid carcinoma cells, AtT-20 corticotrope pituitary cells, RIN2A-19 pancreatic endocrine cells, and NB1 neuroblastoma cells. Northern blot analysis indicated that the AtT-20 cells only expressed precursor convertase 3; in contrast, NB1 cells only expressed precursor convertase 2, whereas the RIN2A-19 and CA-77 cells expressed both enzymes. Despite these differences in expression pattern of precursor convertases the four cell lines were, surprisingly, indistinguishable in respect to their processing of pro-PP and pro-NPY. In all four cell lines, pro-NPY was almost completely converted to NPY, and, in all four cell lines, only around 50% of the PP precursor was converted to PP. The relatively poor processing efficiency of pro-PP was rather similar to the processing efficiency of the endogenously produced precursors in the respective cell lines, pro-calcitonin (CA-77), proopiomelanocortin (AtT-20), proinsulin (RIN2A-19), and pro-vasoactive intestinal polypeptide (NB1). At least in the CA-77 cells, NPY and PP were apparently sorted to the regulated secretory pathway, as upon stimulation with secretagogue the release of the transfected peptides increased in parallel with the endogenously expressed peptide, calcitonin gene-related peptide. Mutagenesis studies showed that on the N-terminal side of the di-basic processing site, the otherwise important difference in structure between PP and NPY, a proline for glutamine in position 34, was not responsible for the difference in processing efficiency. On the C-terminal side of the processing site, the efficient processing of pro-NPY could not be transferred to pro-PP by exchanging the whole C-terminal domains of the precursors. It is concluded that pro-NPY is processed more efficiently than pro-PP in all neuroendocrine cell lines tested independent on their expression of the two main precursor convertases and that mutagenesis data indicate that the structural element responsible for the efficient processing of pro-NPY is not located on the N-terminal side of the dibasic processing site.

Immunohistochemical and chromatographic identification of peptides derived from proneuropeptide Y in the human frontal cortex.

Proneuropeptide Y (proNPY) is posttranslationally processed to NPY(1-36)amide and the C-terminal flanking peptide of NPY (CPON). Antisera directed against the N-terminal part of NPY, CPON, or CysNPY(32-36)amide were used to identify peptide fragments processed from proNPY in biopsies of human frontal cortical specimens obtained from patients who underwent surgical treatment of profound cerebral tumors. Gel filtration and radioimmunoassays of human cortical extracts revealed that the NPY immunoreactivity was found only as NPY(1-36)amide, indicating that all NPY is present in an amidated form. In contrast, no intact proNPY was identified. NPY/CPON-immunoreactive neurons were observed to be nonspiny with long axonal processes mostly orientated longitudinally in the direction of the superficial layers. Bundles of immunoreactive fibers in the underlying white matter were orientated toward superficial layers of the neocortex, indicating a subcortical origin of some NPY/CPON nerve fibers. Axonal terminals were distributed throughout the neocortex, with highest numbers observed in layer I. Some fibers penetrated from the superficial layer I into the overlying pial surface. Many fibers were also observed in proximity to intracortical blood vessels, and some of these fibers originated from the cortical neurons, indicating that NPY could play a role as an intracortical autoregulator of the tonus of cerebral arterioles. Together these results indicate that NPY(1-36)amide and CPON are present in intracortical neurons as two independent molecules and that NPY may be involved in synaptic processes and regulation of blood flow in the human brain.

CHO cells synthesize amidated neuropeptide Y from a C-peptide deleted form of the precursor.

Post-translational processing of peptide precursors producing amidated, biologically active peptides is generally believed to occur only in specially differentiated endocrine or neural cells. Previously it has been shown that endoproteolytic processing of peptide precursors is very inefficient in non-endocrine cells like CHO cells. We have studied the processing of a C-peptide-deleted precursor of neuropeptide Y (NPY) in which the precursor terminates in the sequence Gly-Lys-Arg and does not require any dibasic specific endoproteolytic processing. Following transfection of CHO cells with an expression plasmid encoding this mutated NPY precursor, between 50 and 80 percent of the synthesized NPY was secreted from stable transfectants as authentic amidated NPY as assessed by both a C-terminal amide specific radioimmunoassay and by isoelectric focusing. It is concluded that amidated peptides can be produced in non-endocrine cells provided they are presented with a precursor which does not have to be endoproteolytically processed.

Gastrin releasing peptide in the rat suprachiasmatic nucleus: an immunohistochemical, chromatographic and radioimmunological study.

The localization of gastrin releasing peptide-immunoreactive neurons within the rat suprachiasmatic nucleus was investigated with the indirect peroxidase-antiperoxidase and avidin-biotin techniques. Gastrin releasing peptide-immunoreactive perikarya were found in the ventral part of the suprachiasmatic nucleus, which suggests a direct afferent innervation from the retina or the lateral geniculate, placing the gastrin releasing peptide neurons centrally in the regulation of circadian rhythmicity. Gastrin releasing peptide neurons gave rise to fibers and terminals within the suprachiasmatic nucleus, especially in the ventral part among the immunoreactive and non-immunoreactive neurons. The largest number of gastrin releasing peptide-immunoreactive axons leaving the suprachiasmatic nucleus could be followed in a caudodorsal direction to the subparaventricular zone and the dorsal hypothalamic area. Moreover, minor projections could be traced from the suprachiasmatic nucleus rostrally along the dorsal surface of the optic chiasm to the prechiasmatic area, rostrodorsally into the periventricular area and laterally along the dorsal surface of the optic chiasm and tract. Gel filtration and high performance liquid chromatography profiles of the tissue extracts from the suprachiasmatic area revealed the presence of two GRP-immunoreactive peptides, co-eluting with synthetic gastrin releasing peptide18-27 and gastrin releasing peptide1-27 in almost equivalent concentrations. Gastrin releasing peptide18-27 or gastrin releasing peptide1-27 might therefore play a role in regulation of circadian rhythms in hypothalamic nuclei generated by the suprachiasmatic nucleus.

An immunohistochemical and chromatographic analysis of the distribution and processing of proneuropeptide Y in the rat suprachiasmatic nucleus.

The suprachiasmatic nucleus (SCN) regulates a number of circadian rhythms in mammals. A neuropeptide Y (NPY)-containing pathway from the intergeniculate leaflet of the lateral geniculate to the SCN is considered to carry information of the environmental light-dark cycle. Antisera directed against NPY, Cys-NPY(32-36)amide or the C-terminal extended peptide of proNPY(68-97) (CPON) and avidin-biotin immunohistochemistry were used to define the precise distribution of NPYergic nerve fibers in the SCN, and to compare the location of the various fragments of proNPY in these nerves. Gel chromatography and specific radioimmunoassays were applied to quantify the efficiency of the amidation of NPY, and to study the size of peptides demonstrating NPY- and NPYamide-immunoreactivity in anterior hypothalamic extracts. NPY-, NPYamide-, and CPON-immunoreactive nerve fibers exhibited apparently the same distribution and morphology in the SCN. Immunoreactive fibers were preferentially located in the ventral part of the SCN, but along the rostrocaudal axis of the nucleus, the density and the precise distribution of immunoreactive elements changed. From the rostral third of the SCN to the middle third, the number of immunoreactive fibers increased and their distribution extended in a dorsal and lateral direction. In the caudal part of the SCN, the number of immunoreactive elements decreased and the innervation spread to an even more dorsolateral location. Dorsal aspects of the rostral SCN contained a moderate number of fibers, whereas the dorsomedial quadrant of the caudal 2/3 of the SCN was almost devoid of immunoreactivity.(ABSTRACT TRUNCATED AT 250 WORDS)

Demonstration of neuropeptide Y and its precursor in plasma and follicular fluid.

The present investigation provides three lines of evidence for the presence of a pro-form of neuropeptide Y (NPY) in plasma and follicular fluid. First, by the demonstration of NPY-immunoreactive material of a size corresponding to the estimated mol wt of pro-NPY. Second, an antiserum specific for the C-terminal tyrosine amide of NPY and peptide YY does not react with this material. Third, it was possible to convert the pro-NPY extracted from plasma and follicular fluid using the protease, Endoproteinase-Lys C, to a NPY-immunoreactive form eluting slightly before NPY on a G-50 column. The size of the digested product was consistent with a cleavage of pro-NPY resulting in an immunoreactive species, NPY-Gly-Lys. Pro-NPY was also found in tissue culture media from the human neuroendocrine cell line SH-SY5Y. As in the case of plasma and follicular fluid, another NPY immunoreactive species eluted from a G-50 gel filtration column slightly before synthetic human NPY. Analysis of this material with an antibody directed against the tyrosine amide of NPY in combination with isoelectric focusing revealed that this peak consisted of at least two immunoreactive forms of NPY. In conclusion, at least three different forms of NPY immunoreactivity are likely to be present in plasma, follicular fluid, and cell tissue culture media; pro-NPY, a degradation form of pro-NPY, or a biosynthetic intermediate and NPY.

Partial processing of the neuropeptide Y precursor in transfected CHO cells.

The activation of regulatory peptides by post-translational modification of their biosynthetic precursors is generally thought to occur only in neuroendocrine cells. We have selected clones of Chinese hamster ovary cells, a non-neuroendocrine cell line, which were transfected with a eukaryotic expression vector coding for the precursor for neuropeptide Y. Although the majority of the immunoreactive NPY was found in the form of pro-NPY, some degree of intracellular proteolytic processing of the precursor occurred in all clones. Part of the intracellular NPY immunoreactivity was even correctly amidated. Extracellular degradation of pro-NPY in the tissue culture medium generated immunoreactivity which corresponded in size to NPY. It is concluded that precursor processing can occur in non-neuroendocrine cells both as a biological process within the cells and as apparent processing, degradation in the tissue culture medium.

Neuropeptide Y: presence in sympathetic and parasympathetic innervation of the nasal mucosa.

The occurrence of neuropeptide Y (NPY), vasoactive intestinal polypeptide (VIP) and peptide histidine isoleucine (PHI) in the sympathetic and parasympathetic innervation of the nasal mucosa was studied in various species including man. A dense network of NPY-immunoreactive (IR) fibres was present around arteries and arterioles in the nasal mucosa of all species studied. NPY was also located in nerves around seromucous glands in pig and guinea-pig, but not in rat, cat and man. The NPY-IR glandular innervation corresponded to about 20% of the NPY content of the nasal mucosa as revealed by remaining NPY content determined by radioimmunoassay after sympathectomy. These periglandular NPY-positive fibres had a distribution similar to the VIP-IR and PHI-IR nerves but not to the noradrenergic markers tyrosine hydroxylase (TH) or dopamine-beta-hydroxylase (DBH). The NPY nerves around glands and some perivascular fibres were not influenced by sympathectomy and probably originated in the sphenopalatine ganglion where NPY-IR and VIP-IR ganglion cells were present. The venous sinusoids were innervated by NPY-positive fibres in all species except the cat. Dense NPY and DBH-positive innervation was seen around thick-walled vessels in the pig nasal mucosa; the latter may represent arterio-venous shunts. Double-labelling experiments using TH and DBH, and surgical sympathectomy revealed that the majority of NPY-IR fibres around blood vessels were probably noradrenergic. The NPY-positive perivascular nerves that remained after sympathectomy in the pig nasal mucosa also contained VIP/PHI-IR. The major nasal blood vessels, i.e. sphenopalatine artery and vein, were also densely innervated by NPY-IR fibres of sympathetic origin. Perivascular VIP-IR fibres were present around small arteries, arterioles, venous sinusoids and arterio-venous shunt vessels of the nasal mucosa whereas major nasal vessels received only single VIP-positive nerves. The trigeminal ganglion of the species studied contained only single TH-IR or VIP-IR but no NPY-positive ganglion cells. It is concluded that NPY in the nasal mucosa is mainly present in perivascular nerves of sympathetic origin. In some species, such as pig, glandular and perivascular parasympathetic nerves, probably of VIP/PHI nature, also contain NPY.

Incidental cholecystectomy in the treatment of gynecologic malignancy.

From October 1985 through September 1989, 46 patients with gynecologic malignancy had an incidental cholecystectomy at the time of surgery for their primary disease at the Albany Medical Center Hospital. The mean age was 59 (range 20-87 years). Indications for the gynecologic oncologic operation included endometrial carcinoma in 21 patients, suspected ovarian carcinoma in 17 patients and carcinoma of the cervix in 8 patients. Twenty-three patients (50%) had a preoperative diagnosis of cholelithiasis, and in the remaining 23 patients, the diagnosis of significant gallbladder disease was made intraoperatively. There was only 1 (2.2%) postoperative complication secondary to the cholecystectomy. Prophylactic cholecystectomy accompanying gynecologic cancer surgery can be performed safely and avoids the potential for postoperative cholecystitis and a second operative procedure.

Expression and precursor processing of neuropeptide Y in human pheochromocytoma and neuroblastoma tumors.

The expression of the potent vasoactive peptide neuropeptide Y (NPY) was studied in 16 clinically and/or histologically diagnosed human pheochromocytomas and 3 human neuroblastoma tumors. All tumors contained NPY in concentrations ranging from 21 pmol/g of tissue, similar to that found in normal adrenal tissue, to 91,000 pmol/g (median, 1,700 pmol/g). Three control tumors of Cushing's type did not contain NPY. An almost total proteolytic processing of pro-NPY to normal NPY was observed in the tumors (median, 93%; range, 72-100%). A positive correlation between the processing efficiency and the NPY content was also observed. The small amount of pro-NPY found in the tumors was characterized by "in vitro conversion" with endoproteinase Lys-C. In the tumor extracts, the majority of the NPY immunoreactivity, corresponding in size to the NPY standard, also behaved like synthetic NPY by high performance liquid chromatography and isoelectric focusing. As assessed by both its elution position in isoelectric focusing and its reaction with an antiserum specific for the COOH-terminal amidated sequence, the peptide produced by the tumors was found to be efficiently amidated, a modification which is essential for the biological activity of NPY. It is concluded that although only a subset of chromaffin cells express NPY, a very high number of pheochromocytomas and neuroblastomas produce correctly amidated and thus biologically active NPY in large amounts, and that this is of potential importance for tumor-related cardiovascular symptoms and for autocrine stimulation of tumor cells.

Expression and precursor processing of neuropeptide Y in human and murine neuroblastoma and pheochromocytoma cell lines.

The synthesis and processing of the precursor for neuropeptide Y (NPY) were studied in 16 human and murine neuroendocrine cell lines. Eight of the cell lines, NS-20Y, PC12, LA-N-5, CHP-234, SMS-KCNR, SH-SY5Y, SMS-KCN, and BE(2)-M17, produced sufficient quantities to permit chromatographic characterization of the NPY immunoreactivity. Although the cell lines varied in the amount of NPY they produced, both within and between cell lines, they displayed a relatively constant pattern of posttranslational modifications. In contrast to observations in tumor extracts (M. M. T. O'Hare and T. W. Schwartz, Cancer Res., 49: 7010-7014, 1989), all cell lines studied contained a substantial amount of the intracellular NPY in the form of the unprocessed propeptide, 57% (range, 33-72%) as characterized by both gel filtrations (32 experiments in 8 cell lines) and "in vitro conversion" with endoproteinase Lys-C. In the majority, 4 of 6 cell lines studied, almost all of the NPY, which by size corresponded to the mature 36-amino acid form, was amidated as assessed by isoelectric focusing and by a radioimmunoassay specific for the COOH-terminal amide group of the peptide. Both the propeptide and smaller molecular forms of NPY were secreted from the cell cultures; however, proteolytic degradation in the tissue culture medium prevented a detailed, meaningful characterization of these peptides. It is concluded that many neuroendocrine cell lines, especially those derived from human neuroblastomas, express the NPY gene; the cells display a partly impaired dibasic processing capacity but they generally amidate the products efficiently.

Loss of somatal neuropeptide y immunoreactivity in the rat hippocampus following transient cerebral ischemia.

Adult male Wistar rats were subjected to 20 min of global cerebral ischemia and allowed to survive for 1, 2, 4, or 21 days. The brains were processed for immunocytochemistry and the hippocampal neuropeptide Y (NPY)-immunoreactive (-i) neurons were counted and compared to control values. In order to map out the subregional distribution of ischemic cell loss in the hippocampus, cells were also counted in hematoxylin-eosin (HE)-stained brain sections processed from additional ischemic rats after 21 days survival. Cell counts demonstrated a significant loss of hippocampal NPY-i somata 1-21 days after ischemia. The ischemic loss of somatal NPY-i was in the CAI stratum oriens, the CA1 stratum radiatum, and the CA3(ab) subfield not correlated to hippocampal cell loss. NPY-i fibers were found in all subfields of the hippocampus 1-21 days after ischemia. It is known that the majority (>50%) of hippocampal somatostatin-i (SS) neurons also costore NPY-i and the SS-i neurons in the CA1 and CA3(ab) regions of the hippocampus are preserved following an ischemic insult. The present results showed a 90% ischemic loss of CA1 and CA3 NPY-i somata. Based on these findings, it is concluded that ischemia selectively damaged NPY-i and not SS-i within some surviving hippocampal neurons that co-localized both peptides prior to the ischemia.

Gastrin releasing peptide (GRP) is present in a GRP(1-27) form in anterior pituitary cells of the guinea pig.

Immunohistochemical and chromatographic studies were performed on the guinea pig anterior pituitary gland with an antiserum recognizing an epitope within the gastrin releasing peptide (GRP) carboxyterminal amino acid sequence Val-Gly-His-Leu-Met-NH2. Within the anterior pituitary gland GRP-like immunoreactive cells were identified. The GRP-like immunoreactive cells were distributed heterogenously in the gland, predominantly located in ventral aspects of the anterior pituitary. Intracellularly, the immunoreactivity elements were identified as granula-like structures in the cytoplasma. To further characterize the peptide displaying GRP-like immunoreactivity within the pituitary cells, the GRP-like substances were analyzed by radioimmunoassay and gel filtration chromatography. Using this analytical approach it was determined that the guinea pig pituitary extract contained a peptide with characteristics similar to that of authentic porcine GRP(1-27). Only trace amounts of smaller C-terminal fragments were identified. These results indicate, in contrast to findings in other tissues, the GRP(1-27) is not further degraded into smaller peptide fragments.

The frequency of gastrointestinal endocrine tumours in a well-defined population--Northern Ireland 1970-1985.

The reported incidence of gastrointestinal endocrine tumours is variable. In Northern Ireland circumstances allowing such an assessment are favourable with a central diagnostic laboratory and register established to collect data on tumours from a well-defined population of 1.5 million people. From 1970 to 1985, 368 cases were recorded of which 85 per cent were carcinoid tumours. The annual incidence of carcinoid tumours was 1.3 per 100,000 of the population and the majority occurred in the appendix (61 per cent). No patients presented with the carcinoid syndrome. The annual incidence for other tumours was 0.12 per 100,000 for insulinomas; islet cell tumours of unknown type 0.07; Zollinger-Ellison syndrome 0.05; and multiple endocrine neoplasia (MEN) 0.05. There were two cases of VIPoma, one glucagonoma, one neurotensinoma and one tumour producing ACTH. It is possible that some tumours are more uncommon than others because of difficulty in diagnosis.