Gut: first published as 10.1136/gut.28.11.1417 on 1 November 1987. Downloaded from

Gut, 1987, 28, 1417-1419 Alimentary tract and pancreas in carcinoid syndrome

J,GUSTAFSEN. S BOESBY, F NIELSEN, AND J GIESE Froti th(1 Department of .Surgical Gastroenterology C, Rigshospitalet, Copenhagen 0 and Department of (Clinical Physiology, Glostriup Hospital, Glostriup, Denmark

SUMMARY Bradykinin concentrations in peripheral venous blood were measured in seven patients with carcinoid syndrome. The diagnosis was based on typical symptoms and raised urinary excretion of 5-hydroxy-3-indole acetic acid; the carcinoid tumour was verified histologically. Two patients were flushing constantly and the other patients had flushing attacks two to 10 times daily. Several blood samples were taken at weekly intervals from six of seven patients. During 30 sampling procedures the patients were flushing during sampling in 12 instances. Bradykinin was measured by a sensitive solid phase radioimmunoassay technique. Blood bradykinin concentration was normal in all patients. Bradykinin is unlikely to be the vasoactive mediator of flushing.

The symptoms in carcinoid syndrome may be symptoms of cardiac failure. The diagnosis was caused by the release of several different biologically confirmed histologically and the carcinoid tumour active substances into the bloodstream or activation wais located in the terminal part of the ileum in of vasoiactive substances in the bloodstream.'' aill. The patients presented with metastases at Bradykinin, which is formed in the blood, has laparotomy, six had liver metastases and one http://gut.bmj.com/ traiditionailly been considered to be one of these mesenteric metastases. The median urinary excre- substances and perhaps the most important.' We tion of 5-hydroxy-3-indole acetic acid was 500 have measured bradykinin in peripheral venous mmol/24 h (range 75-1666 mmol/24 h); normal blood in a group of patients with carcinoid syndrome <50 mmol/24 h. using a radioimmunoassay technique and thus Blood samples were taken in five patients five aittempted to clarify the role of bradykinin in times at weekly intervals, in one patient four times at intervals and in one patient only once. The carcinoid syndrome. weekly on October 1, 2021 by guest. Protected copyright. samples were drawn in the morning when the patient Methods was fasting. For a total of 30 sampling procedures, the patients were flushing during sampling in 12 PATI ENTS cases. Three separate blood samples were always Seven patients (four womein, three men, median aige collected consecutively on each day, bradykinin (240 63 years, range 53-74), were included in the study pg) was added to test tube no 2 before sampling. with symptoms from one to 13 yeairs (median six The collection and processing of samples for blood years). All paitients haid typical symptoms of bradykinin measurements have been described in carcinoid syndrome with flushing as their main detail earlier.4 In summary, blood bradykinin was compliaint. Two patients were flushing constaintly and isolated and measured in the following way: blood had diarrhoea, the other patients were flushing two to from an antecubital vein (free.flow samples) was 1t) times daily and had diarrhoea. During the study taken within 1i) s directly into acetone containing none of the patients had any bronchoconstriction or phenanthroline, PolybreneR and EDTA. Lipids were removed by extraction with petroleum ether (40)-60(C). A final purification was made on QAE- Address for correspondicc: clii Gust.sin(is NI1)I).prprtimenti ol Suri ical A-25 at 7-4. The recovery of tracer Giairocn1icrology C( R,ieshiosplalct. ) 111cicd.msxcc IDK-'14(1 Cocpilicagcl 0. Sephadex pH I) miai.rk. amounts of "I-Tyrs-bradykinin was measured in Rccclvicd for piublication ') April 1987 each individual sample. (Overall recovery deter- 1417 Gut: first published as 10.1136/gut.28.11.1417 on 1 November 1987. Downloaded from

1418 Gustaf,vol, BoesbY, Nic[voi, tiii(i GiCS(I minied for 10(0 samples of 6 ml whole blood was pivotal role. Activation of this fa.ctor leads to at 27.8+6.5%, (meanl± SD)).' For unliabelled kalllkrein lnduced formation of bradvkniin. As bradvkinin, the recovery has beenl determiinied as I lailgemia's factor is VerxCasilv activated and as 2 .2±+ 12.5% (mean±l SD, n-=2), when corrected inactivation of fr-ce bradvkinin is fast, coiirect collec- for internal stcandcard recovery)' The senisitivity of the tioni and processing of samples require rapid ilhibi- assay is 3 pg bradykinin/ml blood, and the between tioin and removal of the cizviymes. whliclh generate and assay coefficient of variation is below 1l6. In normal destrov bIraldvkliinil. Results of bradvkinillneilsiere- subjects, venous blood concentration of bradvkinin is menits. whichl have been reported. dlisclose great below 5 pg bradykinin/ml whole blood. vari.ations of tile concentration, and serious qluLestions have beeni raised about the resuilts, whilcil have been Results fouln in StuLdies, where bioatssalys ianid plasma rad.0io- immun11,11o,assays have beenl ulsed.'-' Resuilts witil tilh The concentration of bradykinin in whole blood was currenit assays indicate that the norima.il concentration found to be less than 5 pg/ml in all seven patients, of blood bradkyknin is below 5 pg/ml blood. which is normal. AlthouLgh in 30) samples of peripheral venlotus blood from lpatients with provei carcinoidl syndiromiie Discussion normld blood bradkyknin concentration was fountd to be lcss thain 5 pgrnml wlhole blood. Twelve of the The mrain symptoms of carcinoid syndrome are samples were collected duir-inig fllshinllg. tlushing, diarrhoea, bronchoconstriction and cardiac Our methodological dtata have beenl reported in failure caused by endocardical fibrosis." detail elsewhere.'"' Whole blood is taiken (lirectly In 1953 Lembeck' found that serotonin was into aicetonie withilin I) s. The collectioii tubes aire released frolm carcinoid tumours, and it was believed watshed with inhibitor solution1. These preciUtionIs that flushinig is mediated by serotoniin. It has been prevcnt in il/tm formnation of bradykinin. OfcoLirse, a difficult to ascribe ciarcinoid flush entirely to ain whole blood precipit.atioil with an1 organic solvcit excessive serotonin release, however, because leaids to at conisideraibic loss of . Thcrefore, flushes are not always associated with increased eatch individual tube contains radiolabelled brady- concentrations of circulating serotonin.-" , which serves ais ai recovcry traiccr for the There is evidence that implicaites in the individual sample. Thle recovery is higlhly reproduc- of flush.'-"' In 1 ible, about 28%. The of paithogenesis carcinoid 964 Oates et al staibility 1i-Tyr'-bradykinin http://gut.bmj.com/ found that cutaneous flushes could be produced in throughout the procedurc hais becei documented.' five patients with carcinoid syndrome after a single Experimeilts withi addition of noni-lalbelled brady- intravenous injection of synthetic bradykinin. The kinin before blood sampling hatve slhowIn an overall flushes resembled the attacks which occurred spon- recovery of 80%, ais corrected by meants of recovcry taneously in these patients. Further, the authors tracer data - aid recovery of unlabelled bradykinin found that nine patients with caircinoid syndrome had was controlled in ever-y single blood sampling pro- a significant increase of the kinin concentration cediure in the prcscilt study. We have also cairried out in hepatic venous blood during adrenalin induced across-procedure stability chromatography studies of on October 1, 2021 by guest. Protected copyright. flushes. Their experiments also indicated the exist- endogenous generated bradykinin, which wais pro- ence of an enzyme, which can catalyse the formation duced bh .a delay in addition of acetonie, and we found of a kinin, in metastatic carciinoid tumour tissue. In onie sinigle compound.' 1966 Oates et al' produced cvidenlcc for the release of We conclude that patients with carcinoid syn- bradykinin in carcinoid syndrome. They postulated drome do not haive raised venous blood concenitrat- that carcinoid tumours release the enzyme kallikrein tion of braidykiniin. If braidykinin is the mediator of as a response to stimuli such as adrenalin injection. the peripherall vascular reaiction known als flushing, Kalllikrein is normally found in plasmna in a precursor one would expect to find ralised bradykinin in form. Since these studies it hats been generally peripheral blood. It should be tiakeni into alccount thalt acceptedl that flushing is catused by braidykinin."'' Oates et al' ciarried out their studies 2() years algo. Bronchoconstriction aind perhaps endocardial Since then, there has beeni at treimendous progrcss in fibrosis may also be due to either bradykinin or the methodology of braidykinin meaisuremilents. They serotonin."' stampled large quantities of blood through a small The problems encountered in the measurement of hepatic veini catheter, anid precipitzited blood pro- bradykinin are numerous and difficult. The humran teins with ethanol, whlicll is less efficient than for kallikrein-kinin system is closely linked to the clot- example, acetone. The kinini aictivities in their puri- ting, the complemiienit and the fibriinolytic systems. fied Clualtes were meaIsured using.la bioassay with at with coagulation fiactor XII (I lagemian's factor) in at very low sensitivity ( t()i( ng/mil blood). With our Gut: first published as 10.1136/gut.28.11.1417 on 1 November 1987. Downloaded from

B1?radykinin in (carcinoino syndrome 1419 present knowledge we must conclude that Oates et al biology of the kallikrein-kinin system in health and overestimaited the concentration of circulating brady- disease. In: Pisano JJ, Austen, KF, eds. Fogartv Inter- kinin, mainly because of an in vitro activation of the national Center Proceedings No. 27. Washington: US kaillikrein-kinin system during sample collection. Government Printing Office, 1977: 487-94. 11 Jones RM, Knight D. Severe hypertension and flushing It is now known that carcinoid tumours may in a patient with non-metastatic carcinoid tumour. produce other biologically vasoactive substances A naesthesia 1982; 37: 57-9. such as histamine,2' prostaglandins,' ,'2 12 Herxheimer H. Further observations on the influence of substance K, and eledoisin2' The three latter sub- 5-hydroxytryptamine on bronchial function. J Phvysiol stances belong to the taichykinins. The relationship (Lond) 1953; 122: 49-50. between these substances and the clinical manifesta- 13 Collier HOJ, Holgate JA, Schachter M, Shorley PG. tions is, however, still not known. The bronchoconstrictor action of bradykinin in the We cainnot conclude from this present study that guinca pig. J Pharmacol 1960; 15: 290-7. is without any importance in the develop- 14 Marks C. Carcinoid tumrours: a clinicopathologic stuidl. braidykinin Boston: GK Hall, 1979: 109-1 1. ment of the carcinoid syndrome, but our results 15 Spatz M. Paithogenetic studics of experimentally indicate that braidykinin is unlikely to be the vaso- induced hacirt lcsions and their relattion to the carcinoid active mediator of flushing in these paitients. syndromc. LIab Invest 1964; 13: 288-300. 16 Nielsen F, Darmkjxr Nielsen M, Riasmussen S. References K.appclgaard AM, Gicsc J. Brzadykinin in blood and I Oates IA. Butlcr TC. Pharmacologic aind endocrinc plasma: faicts and fallacies. Acta Med Scand Suppl. 677. aspects ol carcinoid synidromc. In: Garutti S, Shorc PA, 1983; 214: 54-9. cds. Advances in p)IhartnacologV. Vol 5. ILondon: 17 Marceau F, Lussicr A, Regoli 1), Giroud JP. Phairmat- Acadcilic Press Inc, 1967: 109-28. cology of kinins: their rccvance to tissue injury and 2 Robertson JIS, 1elart WS, Andrews 'I'M. The mccha- inflammaition. Geni Pharmnacol 1983; 14: 209-29. nism ol ta1cial flLsh in the carcinoid synidrome. Q J Med 18 Rcgoli I), Barabh J. Pharmacology of briadykiniin and 1962;31: 103-22. related kinins. Pharmnacol Rev 1980; 32: 1-46. 3 (iraharme-Smilth DG. 'Ihc carcinoid syndromc. In: de 19 Shimamrato K, Ando T, Tanaka S, et al. An improved (iroot L. ed. Enidocrinologv. Vol 3. New York: C,runc method lor the determinaition of human blood kiniln and Stratton, 1979: 1721-31. lcvels by sensitive kinin radioimmunoassay. Endocrinol 4 D)armkjxr Nicisenl M. Nielsenl F, Kappelgaard AM, Jpn 1982; 29: 487-94. (iiesc .1. Double-antibody solid phase radioimmuno- 20 Waldenistrim J, Pcrnow B, Silver H. Catsc of matastasis- aissaty lor blood bradykinin. (lin ('him Acta 1982; 125: iig carcinoma (argentalfinoma?) of unknown origin

145-56. showing perculiar red flushing and incrciased amounts of http://gut.bmj.com/ 5 Lembeck F. 5-hydroxytryptamine in carcinoid tumour. histarminc aind 5-hydroxytryptaminc in blood aind urine. Nature 1953; 172: 910-1. Acta Med Scand 1956; 156: 73-83. 6 Levine RJ, Sj6ierdsma A. Pressor amines and the 21 Sandler M. Prostaglandins in amine secreting tumours. carcinoid flush. Ainn Intern Med 1963; 58: 818-28. Lancet 1968; ii: 1053-4. 7 Oates JA, Melmon K, Sj6erdsma A, Gillespie L, Mason 22 HaEikanson R, Bengmark S, Brodin E, et al. Substaince P DT. Release of a kinin peptide in the carcinoid syn- like immunoreactivity in intestinal carcinoid tumours. drome. Lance 1964; i: 5 14-7. In: von Euler US, Pernow B, eds. Substance P. New 8 Oates JA, Pettinger WA, Docter RB. Evidencc for the York: Raiven Press, 1977: 55-8. releasc of bradykinin in carcinoid syndrome. J Clin 23 Norheim 1, Theordorssoni-Norheim E, Oberg K, Brodin on October 1, 2021 by guest. Protected copyright. Invest 1966; 45: 173-8. E, Lundquist G, Rosell S. Antisera ratised against 9 Zeitlin IY, Smith AN. 5-hydroxyindoles and kinins in aind kassinin detect elevated levels of immuno- the carcinoid syndrome and dumping syndrome. Lancet reatctive material in plaisma and tumour tissue from 1966; ii: 986-91. piitients with ciarcinoid tumours. Regul 1984; 9: 1() C'olmann RW, Wong PY, Talamno RC. Chemistry and 245-57.