Letters to the Editors

SIRS, We have read with interest the review article by Jenkins & Berein regarding the use of somatostatin and its long-acting analogue, octreotide, in the treatment of pancreatic diseases.’ We acknowledge their point that much evidence in this field is equivocal, especially with regard to the use of these agents in acute pancreatitis and post-endoscopic retrograde cholangiopancreatography (ERCP) pancreatitis. Many of the difficulties associated with the use of somatostatin and octreotide in these conditions relates to poor definition of the aetiology and pathology of acute pancreatitis. McKay et al. in their review of the field have noted that individual investigators have defined acute pancreatitis in an idiosyncratic manner, leading to a great deal of inter-study variability and general confusion as to the role of somatostatin and octreotide as treatment measures.2 This is highlighted in the aborted study by Sternlieb et aL3 to which Jenkins & Berein refer and which was halted after an apparent increase in post-ERCP complications in a group of patients treated with octreotide. This study was analysed by us at the time4 and was the subject of a long methodological analysis in the volume of the American Journal of Gastroenterology in which the original study a~pea red .~ Once again the problem of defining pancreatitis is at the crux of the confusion here, as Sternlieb and his colleagues employed a liberal definition of post-ERCP pancreatitis. In order to explain the apparent difference between octreotide and somatostatin in the treatment of acute and EKCP-induced pancreatitis, Jenkins & Berein refer to the paper by Binmoeller et al., which deals with the effect of octreotide on the human sphincter of OddL6 To our knowledge this is the only published work relating to this subject, and it found that octreotide increases sphincter of Oddi tone and contractile frequency. In the absence of corroborative evidence these results must stand alone, however we feel that a number of caveats should be entered before using the study’s findings to support Jenkins & Berein’s rather broad hypothesis. In the absence of strong information from human studies, it is usual to look to animal data for clues. However, in the case of the sphincter of Oddi, a major problem arises as it is well recognized that the activity of the sphincter demonstrates a great deal of inter-species variability, thus preventing direct clinical correlation. Moreover, data from animal studies of the effect of somatostatin on sphincter of Oddi function are contradictory, with some reporting stimulation’ while others report inhibition of the sphincter.* We would like to highlight one publication regarding the effect of somatostatin itself on the human sphincter of Oddi. Warzee et al. studied sphincter of Oddi activity in 29 patients, 15 min after the administration of somatostatin (250 pg bolus i.v., then 250 pg/h inf~s ion) .~ After a delay of approximately 4 min, phasic activity (6-10 min) began in the sphincter of Oddi, again lasting for a period of approximately 4 min. This evidence seems to contradict Jenkins & Berein’s contention that octreotide and somatostatin have entirely different activities at the sphincter of Oddi. Given the authors’ hypothesis that treatment with octreotide increases sphincter of Oddi tone and thus causes an increased risk of retention of enzymes, this raises the question of clinical data to support this hypothesis. We believe there are none, as even in large trials such as Binmoeller et al. (245 patients), octreotide treatment is not associated with a higher incidence of pancreatitis than control subjects.1° The most likely reason for this lies in the strong antisecretory effects of octreotide on exocrine pancreas. Numerous studies have detailed the inhibitory effect of octreotide on pancreatic enzyme and fluid secretion both in animals and humans. In 1992, Kemmer et al. reported in this journal that octreotide (even at doses as low as 5 pg/h i.v.) strongly inhibited secretin and ceruletide-stimulated pancreatic enzyme and bicarbonate secreti0n.l’ These findings are in agreement with the work of Gullo and colleagues, who demonstrated (in six healthy volunteers) that octreotide (12.5-50 pg s.c.) significantly inhibited cerulein-stimulated pancreatic enzyme synthesis (as measured by pancreatic amino acid uptake).12 The thrust of these studies is that octreotide can inhibit the production and release of fluid, and more importantly enzymes, from the pancreas. As retention and intracellular activation of pancreatic secretions is thought to be responsible for increased ductal pressure and autodigestion of the gland’s parenchyma, the action of octreotide is likely to