Glucagonomas, VIPomas and somatostatinomas

5 Glucagonomas, VIPomas and Somatostatinomas

STEPHEN R. BLOOM JULIA M. POLAK

Tumour formation in the alimentary system has a distinct pattern. Nonendocrine tumours occur with greatest frequency in the stomach and colon and are not infrequent in the pancreas and oesophagus. In contrast, peptide hormone-secreting endocrine tumours are almost solely located in the pancreas. They can be divided into seven main types. Three of these are the classic gastrinomas and insulinomas and also the so-called ectopic tumours producing such substances as ACTH. In addition there are four newer types, viz glucagonoma, VIPoma, somatostatinoma and PPoma (Bloom and Polak, 1978; Bloom, Polak and WeIbourn, 1979). Of these only the first three have well-recognized clinical syndromes, while the effect of high PP concentrations in the circulation is, as yet, unknown. When examined histologically many of these tumours are found to be composed of several types of APUD cells and secrete more than one of the hormones. Nonetheless, the hormone secreted in greatest quantity dominates the clinical picture and so the clinical syndromes are in practice well distinguished. From a pathological viewpoint, however, it is better to think of these tumours as islet-cell tumours, rather than naming them by their clinical features. There is no doubt that occasional patients can present first with one syndrome and then, either with the passage of time or with the removal of part of the tumour mass, secretion of a second hormonal component comes to dominate the picture, changing the nature of the clinical syndrome. In peptide-secreting tumours, as with other endocrine tumours, the synthetic process of the cell may be considerably disrupted, giving rise to the secretion of high molecular weight forms (Figure 1) whose biological activity may not be identical to that of the smaller moiety (Bloom, Bryant and Adrian, 1978). This is of theoretical interest, as it may give information about the normal pathway of hormone synthesis, and is also of practical relevance as the radioimmunoassayable circulating concentrations of peptide no longer accurately reflect biological activity. Clinics in Endocrinology and Metabolism-Vol. 9, No.2, July 1980.

285

286

STEPHEN R. BLOO M AND JULIA M . POLA K

GLUCAGONOMAS While there have been a number of isolated ease reports in the literature of non-beta cell tumours associated with diabetes and severe clinical wasting , it was not until 1974 that a series of nine cases were gathered together to a llow the description of a 'glucagonoma syndrome' (Mallinson et at. 1( 74). An unexpected, though diagnostically useful. feature of this syndrome is the presence of a necrolytic migratory erythematous rash (Figure 2) (Mallinson and Bloom. 1(78). Indeed, some half of the cases subsequentl y found have come from dermatology clinics. often with the correct specific diagnosis on the referral lett er . Seven main clinical features were orig inally described: severe weight loss . a normochromic normocytic anaemia. diabetes mellitus. a

3C 'J L

0

-

-o

u

0

0

E Q

oj

I

I

'l 5S

]O~

1

U J(

Figure 1. Chromatographic profile of fasting plasma of three patients with the glucagonoma syndrome. The glucagon-like immunoreactivity can be seen to elute from Sephadex GSa in the norm al position for pancreatic glucagon (though in vastly increased amounts) in patient 1 (top panel). In patient 2 (middle panel) a signilkant amount of various higher molecular weight glucagons can be seen. while in the third patient (bottom panel) the vast majority of the glucagon-like immunoreactivity is of greater size than pure pancreatic glucagon. PG Std = pancreatic glucagon standard and Vo = colum n void volume.

GLUCAGONOMAS. VIPOMAS AND SOMATOSTATINOMAS

287

sore red tongue, angular stomatitis, a necrolytic migratory erythematous rash and symptoms referrable to an abdominal pancreatic tumour. Subsequently it has also become apparent that these patients are psychiatrically abnormal and tend to develop overwhelming infection. though both features may be due to the excessive catabolic drive induced by the high circulating glucagon concentrations (Mallinson et al, 1977). In addition a number of

Figure 2. Necrolytic migratory erythematous rash in a patient with the glucagonoma syndrome.

deaths have been described. sometimes during pre-surgical work-up, due to venous thrombosis. There is evidence of increased platelet stickiness in these patients and it may be wise, once the diagnosis is suspected. to initiate aspirin and/or low-dose heparin therapy. Pharmacological experiments in animals and isolated hepatic preparations have shown that low doses of glucagon stimulate glycogenolysis. higher doses stimulate gluconeogenesis and massive doses stimulate ketogenesis and urea formation. In patients with glucagonomas, although diabetes is common,

288

STEPHEN R. BLOOM AND JULIA M. POLAK

ketosis has not yet been described . This may be considered powerful evidence against a major role for glucagon in the development of ketosis in the ordinary diabetic patient. Indeed some glucagonoma patients, even with major weight loss and extensive skin breakdown, still do not have overt clinical diabetes. Presumably in these cases the beta cell is capable of producing enough insulin to override the tendency to excessive gluconeogenesis. The metabolic disturbances secondary to the high gluconeogenesis are probably responsible for many of the features of the syndrome. For example, amino acid concentrations in the circulation are universally low and after tumour removal immediately rise to supranormallevels (Mallinson et al, 1974). The low circulating amino acids may be responsible for the characteristic rash. However, it has also recently been observed that administration of zinc (Figure 3), causes an immediate remission of the rash . This may indicate that the condition is in part the result of zinc deficiency (Mallinson and Bloom, 1978).

VIPOMAS Since the finding, in 1973, that vasoactive intestinal polypeptide (VIP) was produced by pancreatic endocrine tumours associated with severe watery diarrhoea, low gastric acid secretion and hypokalaemia (Bloom. Polak and Pearse, 1973), a considerable number of cases have been collected worldwide. We recently analysed a series of 62 patients with VIP-secreting tumours associated with diarrhoea (Long et al, 1979). Fifty of these patients had pancreatic endocrine tumours secreting VIP. while 10, of whom seven were children, had ganglioneuroblastomas (Figure 4). The patients all had a diarrhoea volume of over 20 mllkg! day, of a highly fluid consistency and free of blood or mucus. Surprisingly, about two-thirds of the patients also complained of abdominal colic and a significant number had an intermittently high faecal fat output. Most patients had significant weight loss, though this was in no case marked. VIP has considerable amino acid sequence similarities to three other well known hormones, secretin. glucagon and gastric inhibitory peptide (GIP). Thus. it was not surprising to find that at operation a torrential secretion of alkaline juice from the pancreas frequently was found. approximately one-fifth of the patients suffered from hyperglycaemia and that gastric acid secretion was also low. VIP, like GIP, is a powerful inhibitor of gastric acid secretion. VIP has been reported to relax the gall bladder and another operative finding that was previously noted in these patients was the presence of a large distended gall bladder. Just under half of the patients had hypercalcaemia. which was partly the result of electrolyte disturbances, partly the result of a high plasma albumin due to dehydration because of the diarrhoea, and in some cases due to coincidental multiple endocrine neoplasia type I and the presence of hyperparathyroidism (see Chapter 4). A fifth of the patients had occasional flushing attacks, presumably due to the vasodilatory action of VIP , which may cause confusion with the carcinoid syndrome. By definition the tumours were mostly composed of VIP-producing cells and the extractable VIP content was extremely high. All patients had

Figure 3. The effect of zinc on the necrolytic migratory erythematous rash: (left) on the night before and (right) three weeks after oral zinc. in a patient with the glucagonoma syndrome.

o

-o

to.I 00

o :s:: > Vl

Z

-l

;!

Vl

a>

o Vl o :s::

> Z

> Vl

:s::

~

-e

Y'

:s:: >

z o

o o

~

r"

290

STEPHEN R. BLOOM AND JULIA M. POLAK

elevated plasma VIP levels, though these showed considerable fluctuations with time . During periods of act ive diarrhoea the VIP level tended to be higher than during periods of remission but the correlation was not close. In healthy young adults the mean plasma VIP concentration is 1.5 pmol/I but levels up to 20 pmol/I are found in some healthy subjects. VIP is released from ischaemic tissue (Modlin. Bloom and Mitchell, 1978a) and seriously ill

gOO

• •

600

.Pancreas oGanq lioneu roblastoma

VIP

(pmol/ll

•• 400

•• •• •••

•••

••

•11• 203±17 ~I.o

200

•• 0111· 11.0 . co

~:::~o

3.4 ±0.4 000

NORMALS (411

•

VIPo.'''AS (62)

Figure 4. The fasting plasma VIP concentration in the first sample received by our laboratory from 62 patients subsequently proven to have VIP-secreting tumours. For comparison. plasma VIP concentrations in 41 control pat ients are shown on the left.

patients often have considerable elevations of circulating VIP. though these rarely exceed 40 pmol/l. In our routine VIP screening radioimmunoassay service only values above 60 pmol/l are reported as suggestive of a VIPoma (see Figure 4). Levels between 20 pmol/1 and 60 pmol/l lead to the request for repeated samples but it is unusual to find that these mildly elevated plasma levels are from patients with a VIPoma. To date all values above 60 pmol/l, other than in patients with terminal shock, have been associated with a VIP·producing tumour. Thus, the measurement of plasma VIP concentration is a very useful clinical screen for the detection of VIPomas.

GLUCAGON OM AS. VIPOMAS AND SOMATOSTATINOMAS

291

Over the last six years a very large number of samples from patients with severe diarrhoea have been received for the measurement of VIP. Eleven patients with purgative abuse, 23 cases of active Crohn's disease and 19 patients with ulcerative colitis with severe diarrhoea and innumerable cases of infective diarrhoea, etc., have all had VIP concentrations within the normal range. In addition 34 patients with carcinoma of the bronchus. 11 patients with medullary carcinoma of the thyroid, 32 patients with the carcinoid syndrome and many patients with adenocarcinoma of the pancreas associated with severe or mild diarrhoea have also been found to have normal plasma VIP concentrations. Thus, it is clear that there are many other mechanisms for the production of diarrhoea which do not involve high circulating VIP levels. The aetiology, for example, of the diarrhoea associated with medullary carcinoma of the thyroid is still unknown. It is our experience to date that all patients with pancreatic tumours, associated with severe and persistent enough diarrhoea to result in significant hypokalaemia, and with normal or reduced gastric acid secretion have had VIP-producing tumours . There have. however. been cases reported in the literature where this has not, apparently, been so. In pursuance of the proposition that VIP-producing tumours are associated with severe diarrhoea. we have investigated the effects of infusing VIP in an experimental pig model. The pigs develop severe watery diarrhoea at a plasma VIP level similar to that seen in human subjects with the VIPoma syndrome (Modlin. Bloom and Mitchel1, 1978b).

SOMATOSTATINOMAS White nearly al1 pancreatic endocrine tumours contain a certain number of D-cells that produce somatostatin in small amounts (Bloom, Polak and West. 1978) it is only recently that patients suffering from hypersomatostatinaemia have been described. The number of cases in the literature is still small but it is clear that following the original description (Ganda et al, 1977; Larsson et al, 1977) many more cases are being detected and it is likely that several larger series will be published in the near future. Somatostatin is a 14 amino acid polypeptide that inhibits numerous endocrine and secretory functions. For example, it completely inhibits the secretion of gastric acid during pentagastrin infusion. but also, and independently, it inhibits the secretion of endogenous gastrin. Thus, it is one of the most powerful known inhibitors of gastric acid. Somatostatin also inhibits pancreatic exocrine secretion both of enzymes and of bicarbonate as wel1 as the endocrine pancreatic secretions of insulin, PP and glucagon (Bloom. 1978). Thus, one would predict that a patient with a high circulating somatostatin concentration would have diabetes, malabsorption and achlorhydria. This was indeed found to be the case. An additional unexpected finding was the presence of gal1 stones. This may be due to the inhibitory action of somatostatin on the choledocus but an additional direct effect on bite salt metabolism cannot be ruled out. These tumours are often very slow growing and we recently had a case where a pancreatic tumour had been found to be unresectable 15 years earlier. The patient had lived in reasonable health, albeit with mild diabetes and steatorrhoea. during this

292

STEPHEN R. BLOOM AND JULIA M. POLAK

period. Post-mortem ex.amin.ation of.the t~mour reveale? massive quantiti~s of somatostatin. Early infusion studies with somatostatin suggested certain dangers. particularly to the haematological system. but the presence of patients in relatively good health with extremely high circulating somatostatin concentrations suggests these earlier fears were not related to somatostatin per se and infusion of purer materials has not produced problems. The measurement of basal somatostatin immunoreactivity in human plasma has proved difficult because of the extremely low concentrations present. In contrast patients with somatostatin-producing tumours have very high circulating levels and diagnosis is therefore not difficult . Much of the circulating immunoreactive material is. as with other tumours, of high molecular weight. The biological activity of this material is unknown . PPOMAS In 1972 Chance. in the Eli Lilly Laboratories in Indianapolis. extracted from insulin a single major protein impurity. which he termed pancreatic polypeptide (PP) (Lin and Chance, 1972). This material was subsequently found to be present in the circulation and to be produced and stored in endocrine cells in the pancreas (Heitz et aI, 1976). Pharmacological studies have shown that PP inhibits gall bladder contraction and pancreatic enzyme secretion (Greenberg et aI, 1978). However, this effect is weak and can easily be over-ridden . PP is frequently found in pancreatic endocrine tumours (Polak et al, 1976); approximately SO per cent of them are found to secrete PP in addition to their main product (producing the clinical syndrome) (Table 1). The measurement of plasma PP may. therefore, prove of

Table 1. Plasma PP in patients with pancreatic tumours

Tumour PPoma VIPoma Glucagonoma Gastrinoma Insulinoma Adenocarcinoma

Total number studied 1 49

14 60 17 16

Number with fasting PP elevated (>300 pmol/l)

Per cent positive

1 31 7 IS 4

100

o

o

63

SO 2S 24

diagnostic help in certain circumstances. For example, if VIP is found to be very high in a child with watery diarrhoea, the presence. in addition, of very high PP values suggests a pancreatic tumour rather than a ganglioneuroblastoma. In the latter tumours PP values have never been found to be elevated. To date no clinical difference has been noted between those cases with extremely high PP values and those without PP elevation who suffer from the same hormone excess. Thus. the high PP levels are clinically silent.

GLUCAGONOMAS. VIPOMAS AND SOMATOSTATINOMAS

293

DIAGNOSIS Over the last few years the incidence of pancreatic endocrine tumours has increased very greatly. It is most likely that this is not a real phenomenon but is due to enhanced awareness by clinicians of the nature of the clinical syndromes produced. However, it is still true that many cases are found only in the most advanced stages, or at post-mortem. Knowledge of the- precise clinical picture and also of the dramatic clinical improvements that can be obtained by tumour removal should lead to enhanced clinical awareness and thus earlier requests for measurement of the relevant hormone.

RADIOIMMUNOASSAYS While the radioimmunoassay of gut hormones is both relatively cheap and highly specific, there are, nonetheless, a number of problems. Firstly such assays depend on exact and precise quantitative binding between an antibody and an antigen. Any factor which changes, even slightly, the tertiary structure of the antibody will necessarily result in erroneous answers. There are several such factors in plasma which may do this. A second problem is that antibody combines with a specific amino acid sequence. This amino acid sequence of a peptide may not go along with biological activity. Thus, for example, in renal failure, where hormone fragments, normally cleared from the circulation by the kidney, are present in high concentrations, laboratories will report high hormone values that are of no clinical significance. Over the last few years there have been considerable improvements in basic laboratory techniques with the development of proper quality controls. In addition antibodies of greater sensitivity have been raised so the chance of nonspecific interference is lessened. Antibodies have also been produced which react specifically with the biologically active site ofthe hormone and therefore tend to give information more closely related to its biological activity. Scientific developments in the field of radioimmunoassay can be expected to continue and, in particular, to lead to automation. This in turn will make assay services more widely available to the peripheral clinician. At the present time VIPomas, somatostatinomas and glucagonomas are diagnosed late in their clinical course and hormone concentrations are extremely elevated. Thus the measurement of even a single fasting sample is sufficient to suggest the diagnosis. It can be predicted that in future such patients will be investigated at an earlier stage of the disease when, as is already the case with gastrinomas and insulinomas, diagnostic problems will be greater. There have been no specific stimulation or suppression tests yet developed for any of these three newer syndromes, though it is clear that some such test will be required. One feature, mentioned earlier, is the presence of abnormal hormone forms being produced by tumours. Detection of these, at the present time, requires laborious chromatography of each plasma sample and is therefore not feasible on a large scale. Nonetheless, it is our experience that almost every tumour can be distinguished from 'physiological' hypersecretion states by the presence of these abnormal hormone forms.

294

STEPHEN R. BLOOM AND JULIA M. POLAK

LOCALIZATION Localization of an endocrine tumour can be performed by all normal oncology techniques. Endocrine tumours, however, are distinguished by the fact that the clinical problem results from hormone secretion rather than tumour bulk. Thus even an extremely small tumour can prove fatal. Many insuunomas. for example. may be no larger than half a centimetre in diameter and yet cause fatal hypoglycaemia. Special techniques are therefore required to localize these tiny tumours. Considerable interest has been shown in the techniques of venous sampling to localize 'hot spots' of hormone production (Ingemansson, Lunderquist and Holst, 1976). These are best performed on a background of suppression of normal hormone release (for example, hypoglycaemia for insulinomas). Unfortunately a number of problems have arisen. Firstly. the pancreatic tumours drain into the portal circulation and the only feasible method of catheterization is by transhepatic portal venous sampling (Kingham et aI, 1978). This procedure is relatively difficult to perform and is associated with risk of hepatic haemorrhage, biliary leakage and perforation of the bowel. Furthermore, as tumours may be multiple and often have aberrant venous drainage, localization may not be sufficiently precise to be of assistance to the surgeon. In our experience the techniques of highly selective arterial catheterization with background subtraction films of the resulting angiogram are more reliable and have certainly proven less hazardous (Allison, Modlin and Jenkins, 1977). To date we have not failed to localize a tumour by the latter technique.

TREATMENT Successful complete surgical resection of an endocrine tumour is clearly the optimal goal (Bloom, Polak and Welbourn, 1979). Unfortunately, tumours may be multiple and it also seems likely that in some cases a generalized 'field change' leads to subsequent development of new primaries. Approximately half the cases of glucagonomas, VIPomas and somatostatinomas are diagnosed after obvious metastases to local lymph nodes or the liver have already occurred. In these cases removal of tumour bulk may be helpful in prov iding the patient with relief of symptoms. If, at laparotomy, it is not possible to remove tumour mass , then a useful new technique is selective arterial embolization (Allison, Modlin and Jenkins , 1977). This is only applicable to hepatic metastases, however. The cytotoxic agent, streptozotocin, originally developed for treatment of insulinomas, has proved dramatically successful in the treatment of VIPomas (Figure 5), often producing remission lasting many years (Gagel et al, 1976). Its effect on gastrinomas is far less successful and to date there is an inadequate experience with somatostatinomas. Symptomatic treatment is also possible with a number of different agents. Our group has been particularly interested in the use of the somatostatin analogue Des' 2 • S 12 Il D Try" somatostatin, which, following subcutaneous injection, can suppress hormone production from tumours for up to 24 hours (Long et al , 1979) (Figure 6).

295

GLUCAGONOMAS. VIPOMAS AND SOMATOSTATINOMAS St,eptozoloc,n. 500 mg/sq m, 5 doy couru

+

O

Cl LU

::t

~

15

•

+

Prednisone. 40-60 mg/doy

o

600

500

400 E .....

:r'300 1I:

0-"

> ZOO

_

Upper Ronge 01

~

100

~'."~J

-----------o-o~o

July 73

Ocl 73

Jon

Apr

July

OCI

Jon

Apr

74

74

74

74

75

75

TIME

Figure 5. The effect of streptozotocin on the diarrhoea production and plasma VIP concentration in a patient with an unresectable VIPoma. (I pmol/I VIP = 3.3 pg/ml.)

.-..1 l_. Q ;,:,.e --.J

100

o~

r-

0

\

I I

/~ o

'" ro '"

.0

\

of-

c

\

50

D-""""'0

\\\

0

C"

'"

•

oJ ::J

<.:>

o

o -24

o

/

-"""0

/

r

/ I

0

I

' " '~l/ I a I

0

'l'•.,

I

~~ '_. __. - - - - --~----.

6

,'-

,,"

~

12

hours Figure 6. The effect of administering 5 mg Des' I • I II Il D Try' somatostatin analogue subcutaneously to two patients with the glucagonoma syndrome. The plasma glucagon concentration is expressed as a percentage of the value at zero.

296

STEPHEN R. BLOOM AND JULIA M. POLAK

CONCLUSIONS The three new endocrine pancreatic tumour syndromes dealt with in this chapter - glucagonomas, VIPomas and somatostatinomas - are not common. Nonetheless, the patients are potentially curable by tumour resection and therefore wider knowledge of the clinical picture is of considerable importance. It is possible that there are still further, presently unrecognized, clinical syndromes waiting in the wings and studies are currently under way to try to ascertain the effects of elevated PP. Early tumour diagnosis depends firstly on clinical acumen but the easy availability of a reliable radioimmunoassay service is of considerable importance as this allows the physician to screen likely patients and so detect cases at an early stage. Tumour localization is still a major problem and requires expert radiological assistance. The dramatic effectiveness of the cytotoxic agent streptozotocin illustrates the great potential of chemotherapy and it may be expected that further drugs of this nature will be discovered, especially as it is now possible to establish isolated tumour strains for mass drug screening. These tumours have shown the effect of long continued peptide elevation and given valuable insight into the physiological role of the respective regulatory peptides.

REFERENCES

Allison, D. J.• Modlin, I. M. & Jenkins, W. J. (1977) Treatment of carcinoid liver metastases by hepatic-artery embolisation. Lancet. II, 1323-1325. Bloom. S. R. (1978) Somatostatin and gut. Gastroenterology. 75, 145-147. Bloom, S. R. & Polak. J. M. (1978) The glucagonoma syndrome. In Advances in Experimental Medicine and Biology (Ed.) Grossman. M., Speranza, V., Basso, N. & Lezoche, E. pp. 183-194. NewYork: Plenum Press. Bloom, S. R.. Bryant, M. G. & Adrian, T. E. (1978) Multiple hormone forms in PPomas, VIPomas and glucagonomas. Scandinavian Journal of Gastroenterology. 13,24. Bloom, S. R.. Polak. J. M. & Pearse, A. G. E. (1973) Vasoactive intestinal peptide and watery diarrhoea syndrome. Lancet. 11,14-16. Bloom, S. R., Polak, J. M. & West, A. M. (1978) Somatostatin content of pancreatic endocrine tumours. Proceedings of International Symposium on Somatostatin. Metabolism. 27, 1235-1238. Bloom. S. R., Polak. J. M. & Welbourn, R. B. (1979) Pancreatic apudomas. World Journal of Surgery. 3,587-595. Gagel, R. F., Costanza, M. E.. De Lellis, R. A., Norton. R. A., Bloom, S. R., Miller, H. H., Ucci, A. & Nathanson. L. (1976) Streptozotocin-treated Verner-Morrison syndrome. Plasma vasoactive intestinal peptide and tumour responses. Archives of International Medicine. 136,1429. Ganda, O. P.• Weir, G. C., Soeldner, 1. S., Legg, M. A., Chick, W. L.. Patel, Y. C., Ebeid, A. M., Gabbay, K. H. & Reichlin, S. (1977) Somatostatinoma - a somatostatin-containing tumour of the endocrine pancreas. New England Journal of Medicine. 297, 1352-1357. Greenberg, G. R., McCloy. R. F., Adrian, T. E., Chadwick, V. S.• Baron. J. H. & Bloom, S. R. (1978) Inhibition of pancreas and gall bladder by pancreatic polypeptide. Lancet. II, 12801282. Heitz, P., Polak, J. M., Bloom, S. R. & Pearse, A. G. E. (1976) Identification of the D,-cell as the source of human pancreatic polypeptide (HPP). Gut. 17,755. Ingemansson, S.• Lunderquist, A. & Holst, J. (1976) Selective catheterization of the pancreatic vein for radioimmunoassay in glucagon-secreting carcinoma of the pancreas. Radiology. 119,555·556.

GLUCAGONOMAS. VIPOMAS AND SOMATOSTATINOMAS

297

Kingham, J. G. C.• Dick. R.• Bloom, S. R. & Frankel, R. J. (1978) VIPoma: localisation by percutaneous transhepatic portal venous sampling. British Medical Journal. II, 1682-1683. Larsson. L-\., Hirsch. M. A.• Holst. J. J.. Ingemansson, S.. Kuhl, C.. Lindkaer Jensen, S., Lundqvist. G .• Rehfeld, J. F. & Schwartz. T. W. (1977) Pancreatic somatostatinoma. Clinical features and physiological implications. Lancet. 1,666. Lin. T. M. & Chance. R. E. (1972) Spectrum of gastrointestinal actions on a new bovine pancreatic polypeptide (BPP). Gastroenterology. 62,852. Long. R. G .• Barnes. A. J .• Adrian. T. E.• Mallinson. C. N., Brown, M. R.. Vale. W.• Rivier, J. E.• Christofides, N. D. & Bloom. S. R. (l979a) Suppression of pancreatic endocrine tumour secretion by long-acting somatostatin analogue. Lancet. 11,764-767. Long. R. G .. Mitchell. S. J.. Bryant. M. G., Polak. J. M. & Bloom. S. R. (l979b) Clinicopathological study of pancreatic and neural VIPomas. Gut. A934. Mallinson. C. N. & Bloom. S. R. (1978) The hyperglycemic. cutaneous syndrome: pancreatic glucagonoma. In Surgical Endocrinology: Clinical Syndromes (Ed.) Friesen. S. R. pp. 171202. Philadelphia, Toronto: J. B. Lippincott. Mallinson. C. N.. Bloom. S. R.. Warin, A. P.. Salmon. P. R. & Cox. B. (1974) A glucagonoma syndrome. Lancet. II, I. Mallinson. C. N.. Adrian. T. E.• Hanley, J., Bryant, M. & Bloom. S. R. (1977) Metabolic and clinical responses in patients with pancreatic glucagonomas. Irish JOlmUlI 0/ Medical Science. 146,37. Modlin. \. M.• Bloom. S. R. & Mitchell, S. J. (l978a) Plasma vasoactive intestinal polypeptide (VIP) levels and intestinal ischemia. Experientia, 34,535-536. Modlin. \. M.• Bloom. S. R. & Mitchell. S. J. (l978b) Experimental evidence for vasoactive intestinal peptide as the cause of the watery diarrhoea syndrome. Gastroenterotogy. 75, 1051-1054. Polak. J. M.. Bloom. S. R.. Adrian. T. E.• Heitz, P.• Bryant. M. G. & Pearse. A. G. E. (1976) Pancreatic polypeptide in insulinomas, gastrinomas, VIPomas and glucagonomas. Lancet. 1,328-330.