- Email: [email protected]
Role of octreotide in the prevention of postoperative complications following pancreatic resection
Role of Octreotide in the Prevention of Postoperative Complications Following Pancreatic Resection Markus Btichler, MD, Helmut Friess, MD, Istvan Klempa, MD, Peter Hermanek, MD, Udo Sulkowski, MD, Heinz Becker, MD, Anton Schafmayer, MD, ho Baca, MD, Dietmar Lorenz, MD, Richard Meister, MD, Bernd Kremer, MD, Peter Wagner, MD, Jens Witte, MD, Ernst Ludwig Zurmayer, MD, Hans-Detlev Saeger, MD, Bernd Rieck, MD, Peter Dollinger, MD, Karl Glaser, MD, Reinhard Teichmann, MD, Jochen Konradt, MD, Wilhelm Gaus, Phn, Hans-Joachim Dennler, MD, Dieter Welzel, MD, Hans G. Beger, MD
Though morbidity and mortality rates following pancreatic resection have improved in recent years, they are still around 35% and 5%, respectively. Typical complications, such as pancreatic fistula, abscess, and subsequent sepsis, are chiefly associated with exocrine pancreatic secretion. In order to clarify whether the perioperative inhibition of exocrine pancreatic secretion prevents complications, we assessed the efficacy of octreotide, a long-acting somatostatin analogue. We conducted a randomized, double-blind, placebo-controlled, multicenter trial in 246 patients undergoing major elective pancreatic surgery. Patients were stratified into a high-risk stratum (limited to patients with pancreatic and periampullary tumors) or low-risk stratum (patients with chronic pancreatitis) . Patients received octreotide (3 X 100 pg) or placebo subcutaneously for 7 days perioperatively. Eleven complications were defined: death, leakage of anastomosis, pancreatic fistula, abscess, fluid collection, shock, sepsis, bleeding, pulmonary insufficiency, renal insufficiency, and postoperative pancreatitis. Two hundred patients underwent pancreatic head resection, 31 patients underwent left resection, and 15 patients had other procedures. The overall mortality rate within 90 days was 4.5%, with 3.2% in the octreotide group and 5.8% in the placebo group. The complication rate was 32% in the patients receiving octreotide (40 of 125 patients) and 55% in patients receiving placebo (67 of 121 patients) (p
A complete list of all affiliations appears on page 129. Requests for reprints should be addressed to Hans G. Beger, MD, FACS, Department of General Surgery, University of Ulm, Steinhoevelstrasse 9, D-7900 Ulm, Germany. Presented at the 32nd Annual Meeting of the Society for Surgery of the Alimentary Tract, New Orleans, Louisiana, May 20-22, 1991.
The perioperative application of octreotide reduces the occurrence of typical postoperative complications after pancreatic resection, particularly in patients with tumors.
hough the mortality rate following pancreatic resecT tion has considerably improved in recent years, it remains between 3% and 10% [I-6]. Complications including medical and surgical morbidity occur in 30% to 40% of the patients [ 1,3,6-181. Major complications after pancreatic resection, some of which can be fatal, such as pancreatic fistula, intra-abdominal fluid collection, ab scess, and subsequent sepsis, are chiefly associated with exocrine pancreatic secretion. The concept of inhibiting exocrine pancreatic secretion to prevent postoperative complications was originated in 1979 [ 191. In an open trial, a continuous intravenous infusion (250 pg/h) of somatostatin was administered to patients perioperatively after Whipple’s resection. The authors of the study reported a reduced complication rate. The human hormone somatostatin has been demonstrated to strongly inhibit basal and stimulated exocrine pancreatic secretion [20-221. Recently, the synthetic somatostatin analogue, octreotide (SMS 201995), has become clinically available, and it is more potent in inhibiting pancreatic enzyme secretion [23,24] and has a much longer half-life [25,26] than does the native hormone. In a multicenter controlled and randomized trial, we analyzed the value of the perioperative application (7 days) of octreotide in patients undergoing major elective pancreatic surgery. PATIENTS
AND METHODS
Between October 1989 and July 1990, 322 patients were recruited from 18 surgical departments in Germany (n = 17) and Austria (n = 1). The study was approved by the ethics committees of each of the participating hospitals. Patients with pancreatic or periampullary tumors and those suffering from chronic pancreatitis who were suitable for elective pancreatic surgery (resection) were enrolled in the study after giving written, informed consent. Patients were excluded who were undergoing emergency surgery of the pancreas (acute pancreatitis, pancreatic trauma), total pancreatectomy, or pancreatic transplantation, or who were receiving only an elective pancreatic-cyst anastomosis. On the day before the operation, the patients who satisfied the inclusion criteria were randomly assigned to receive a 7-day treatment of
THE AMERICAN JOURNAL OF SURGERY
VOLUME 163 JANUARY 1992
125
BtiCHLER
ET AL
either 3 X 100 pg octreotide (1 mL volume) subcutaneously (every 8 hours) or 21 subcutaneous placebo injections (1 mL each) in a double-blinded fashion. Treatment was initiated at 7:00 AM on the day of operation, in each patient at least 1 hour before the laparotomy. The dosage of 3 X 100 pg subcutaneously per day was chosen after a pilot trial [27] had shown that this dosage effects maximum inhibition of pancreatic enzyme secretion in humans. The patients were stratified into two groups: those with pancreatic or periampullary tumors (high-risk stratum) and those suffering from chronic pancreatitis (low-risk stratum). Randomization was carried out separately for each hospital and for each stratum. Each patient received perioperative intravenous antibiotic prophylaxis for 24 to 48 hours using bactericidal drugs effective against gram-negative bacteria. The patients were treated in an intensive care unit for at least 3 days postoperatively. Intravenous crystalloid solutions were administered (2.5 to 3.0 L/day) according to clinical requirements and were supplemented by plasma or blood components in order to maintain a positive central venous pressure and a blood hematocrit greater than or equal to 30%. The patients were permitted nothing orally for at least 3 postoperative days. Definition of postoperative complications: The study protocol defined 11 typical complications occurring after pancreatic resection and stipulated that each patient be followed for 90 days postoperatively. The complications were defined as follows: (1) Death (within 90 days postoperatively). (2) Leakage of pancreatic, biliary, or intestinal anastomosis as determined by radiograph or intraoperative (relaparotomy) findings. (3) Pancreatic fistula was defined by a concentration of amylase and lipase in the drainage fluid more than 3 days postoperatively of more than three times the serum concentration and a drainage volume of more than 10 mL/24 hours at the same time. The serum and drainage fluid amylase and lipase concentration was determined on postoperative days 1, 3,4, 5, and 7 and twice a week thereafter. (4) An intra-abdominal abscess was defined as a pus collection or an infected fluid collection revealed by ultrasound or computed tomography (CT) guided needle aspiration and microbiological culture or demonstrated by relaparotomy. (5) An intra-abdominal fluid collection (sterile) was defined as a collection of fluid of at least 5 X 5 cm in diameter in ultrasound or CT that did not qualify as an abscess. All patients were screened for a fluid collection by means of ultrasound or CT on day 4 and day 8 postoperatively. (6) Cardiocirculatory shock was defined as a systolic blood pressure of less than 80 mm Hg for more than 20 minutes (later than 24 hours postoperatively) that required continuous volume replacement and/or treatment with catecholamines. (7) Sepsis was defined as the presence of at least four of the following five criteria: (A) bacterial-positive blood culture, (B) rectal temperature greater than 38.5% (C)
leukocytosis greater than 12,000/mm3 or leukopenia less than 3,000/mm3, (D) thrombocytopenia less than 100,000/mm3, and (E) metabolic acidosis (base excess greater than -4). Blood cultures were taken daily in patients with rectal temperature of higher than 38.5’C. Blood leukocyte and thrombocyte determination and a blood gas analysis were performed on postoperative days 1, 3, 4, 5, and 7 and twice a week thereafter. (8) Pulmonary insufficiency was defined as an arterial PO2 of less than 60 mm Hg, despite oxygen application (4 L/m) via the nose, or when the patient required prolonged postoperative mechanical ventilation (greater than 24 hours after the end of surgery) or reintubation. (9) Renal insufficiency was defined as an increase in serum creatinine of more than twice the preoperative value. Patients whose serum creatinine was increased preoperatively (greater than 120 pmol/L) were not included in the analysis of this criterion. Serum creatinine was monitored on postoperative days 1, 3, 4, 5, and 7 and twice a week thereafter. (10) Bleeding was defined as the replacement of more than 3 units of blood more than 24 hours after the end of surgery or the indication for relaparotomy due to intraabdominal or intra-intestinal bleeding. Blood hematocrit was monitored daily on postoperative days 1,3,4,5, and 7 and twice a week thereafter. (11) Postoperative acute pancreatitis was defined as a more than threefold increase in serum amylase or lipase more than 3 days after surgery, accompanied by morphological signs of acute pancreatitis in a contrast-enhanced CT scan. CT was mandatory in patients with a threefold increase in serum amylase or lipase on postoperative day 4 or after. Statistical analysis: During the design of the study, we anticipated a rate of 30% to 40% patients with one or more complications in the placebo group. A power calculation for a two-tailed significance level of 0.05 indicated that it would be reasonable to recruit about 100 patients for each group. This number of patients enabled the study to detect a reduction of complication rate from 30% to 40% in the placebo group to 15% to 20% in the octreotide group with a power of 90%. The hypothesis testing was done with the Wilcoxon two-sample rank sum test for qualitative outcome variables, with the chi-square test for quantitative outcome variables. For testing small subgroups, the chi-square test was corrected for discontinuity as proposed by Williams [28]. The hypotheses for the comparison of the complication rates between the placebo and octreotide groups were formulated a priori, making this test strictly confirmatory.
126
163
THE AMERICAN
JOURNAL
OF SURGERY
VOLUME
RESULTS
On the day before the operation, 322 patients were randomized and received either a primary subcutaneous injection of placebo or 100 pg of octreotide in the morning of the day of surgery. During the operation, 246 of the 322 patients (76.4%) were found to be operable (resectable), according to the conditions stipulated in the study protocol. Seventy-six subjects were withdrawn from the study, and the injection therapy was discontinued beJANUARY
1992
DCTREOTIDE TREATMENT FOLLOWING PANCREATIC RESECTION
TABLE
TABLE
I
II Mortality and Complications
Comparablllty of Groups
Age (median, range) Sex (m/f) Type of disease Pancreatic cancer Periampullary cancer Endocrine tumor Chronic pancreatitis Others
Octreotide (n = 125)
Placebo (n = 121)
51 (19-75) 92133
52 (20-80) 82139
37 15 5 82 8
Complication rate (patients with 1 or more complications) Number of complications (Total 208)
34 25 4 50 8
73 26
79 22
18 3 2 3
13 5 1 1
4 (3.2%)
7 (5.8%)
40 (32.0%) p < 0.005 67 (55.4%) 126
82
III
Number and Type of Complications Octreotide (n = 125)
cause pancreatic resection was not possible due to advanced tumor disease (n = 73) or because only a pancreatic-cyst anastomosis was carried out (n = 3). In three patients (two from the octreotide group, one from the placebo group), therapy was discontinued after 2 and 4 days, respectively, due to evidence of side effects. The three patients were retained in the trial. A total of 246 patients were included in the final evaluation, with 125 subjects assigned to the octreotide group and 12 1 to the placebo group. The two groups were well matched with regard to age, sex, type of underlying disease, and type of surgery performed (Table I). F’reoperative morbidity (heart, lung, liver, kidney, diabetes mellitus) were comparable between the two groups. Of the 246 patients, 139 were in the high-risk stratum (tumors) and 107 in the low-risk stratum (chronic pancreatitis). Pancreaticoduodenectomy (Whipple’s operation [29]) was performed in 152 patients, a duodenumpreserving pancreatic head resection [30] in 48 patients, a pancreatic left resection in 3 1 patients, pancreaticojejunostomy in 8 patients, enucleation for endocrine tumor in 3 patients, and atypical pancreatic resection in 4 patients (Table I). The overall mortality within 90 days of surgery was 4.5% (11 of 246 patients). -Four patients (3.2%) of the octreotide group and 7 patients (5.8%) of the placebo group died (Table II). Among the high-risk patients (139 patients), the mortality rates were 2.9% and 9.9% in octreotide and placebo groups, respectively. Forty patients (32%) of the octreotide group and 67 (55.4%) of the placebo group suffered one or more postoperative complications as defined in the study protocol (Table II). This difference was significant at a level of p <0.005. The total complication rate was 43.5% (107 of 246 patients). In these 107 patients, a total of 208 complications were recorded, 82 and 126 in the octreotide and placebo groups, respectively (Table II). The most frequent complication was pancreatic fwtula (27%), followed by pulTHE AMERICAN
Placebo (n = 121)
Mortality
TABLE Type of operation Whipple’s Duodenum-preserving pancreatic head resection Left resection Pancreaticojejunostomy Enucleation Others
Octreotide (n = 125)
4 6 22 8 8 6 3 12 1 12 0 82
Death Leakage of anastomosis Pancreatic fistula Abscess Fluid collection Shock Sepsis Pulmonary insufficiency Renal insufficiency Bleeding Postoperative pancreatiiis Total complications
TABLE
Placebo (n = 121) 7 7 46 12 9 7 6 16 2 10 4 126
IV
Number of Patients With One or More Complications in High-Risk and Low-Risk Strata Octreotide
Placebo
No. of high-risk* patients (n = 139)
68
71
No. of high-risk patients with one or more complications
26 (38%)
No. of low-riskf patients (n = 107)
57
No. of low-risk patients with one or more complications
14 (25%)
‘Patients with pancreatic or periampullary tPatients with chronic pancreatitis.
p co.01
46 (65%) 50
NS
21 (42%)
tumors.
monary insufficiency (1 l%), bleeding (9%), abscess (8%), fluid collection (7%), anastomosis leakage (5%) shock (5%), death (4.5%), sepsis (4%), pancreatitis (2%), and renal insufficiency ( 1%). The frequency of death, pancreatic listula, abscess, sepsis, pulmonary insufficiency, and postoperative pancreatitis was lower in the patients in octreotide group than in the placebo group (Table III). The rates of complications among the high-risk patients were 38% for the octreotide group and 65% for the placebo group (p
JOURNAL
OF SURGERY
VOLUME
163
JANUARY
1992
127
BikHLER
TABLE
ET AL
V
Number and Type of Complications in Patients in the High-Risk Stratum (Tumors) Octreotide (n = 66)
Placebo (n = 71)
2 4 16 4 5 4 1 7 0 7 0
7 6 29 IO 7 7 5 12 2 9 4 96
Death Leakage of anastomosis Pancreatic fistula Abscess Fluid collection Shock Sepsis Pulmonary insufficiency Renal insufficiency Bleeding Postoperative pancreatitis Total number of complications
4
after the operation) and 1 (lipase and amylase). This was lower in the patients receiving octreotide than in those in the placebo group (Figure 1). The mean hospitalization time of the 246 evaluable patients was 24.1 f 1.2 days. The patients in the octreotide and placebo groups were hospitalized for 22.1 f 1.5 and 26.2 f 1.9 days, respectively. In patients in the highrisk stratum who had a mean hospitalization of 32.2 f 2.6 days, those in the octreotide group remained in the hospital 29.0 f 2.4 days and in the placebo group 35.3 f 2.9 days. Fifty-six patients (31 octreotide, 25 placebo) experienced pain during the subcutaneous injection at the injection site, which did not require discontinuation of the treatment. Two patients (both in the octreotide group) suffered hot flashes, and the treatment was stopped on day 2 and 4, respectively. One patient (in the placebo group) had an inflammatory reaction at the injection site that led to discontinuation of the treatment on day 4. Two other patients suffered from itching and temporary migrant exanthema, which did not necessitate discontinuation of the treatment. No further side effects were registered.
evidenced by the number and type of complications (Table V). Table V also shows a difference in the occurrence of death, pancreatic fistula, abscess, shock, sepsis, pulmonary insufficiency, and postoperative pancreatitis. Serum pancreatic enzyme values showed a significant difference on postoperative days 0 (lipase) (in the evening
COMMENTS Somatostatin, a tetradecapeptide originally discovered in hypothalamic tissue, is a powerful inhibitor of baseline and stimulated exocrine pancreatic secretion [20-221. Its action on the pancreas is mediated via the inhibition of secretagogue gastrointestinal hormones and by direct pathways via somatostatin receptors at acinus cells [31]. Octreotide (SMS 201-995), an octapeptide analogue of somatostatin, was synthesized to have morespecific, more-potent, and longer-acting inhibitory effects than native somatostatin [25,26j. The longer half-life of octreotide (100 min), in particular, in comparison with somatostatin 14 (2 min) has made this compound advantageous for clinical application. In a pilot trial, we were able to show that 100 pg octreotide administered subcutaneously inhibited the stimulated secretion of amylase, trypsin, and chymotrypsin by 84%,76%, and 77%, respectively [27]. Our data confirm those of earlier reports that showed a maximum inhibition of pancreatic protease and amylase secretion with comparable dosages of octreotide [23,24]. In pancreatic surgery, the pancreatic remnant and its anastomosis with stomach or small bowel constitutes the major risk factor. This problem is caused by a continuous exocrine pancreatic secretion (including the secretion of harmful proteases and lipolytic enzymes) of the pancreatic stump and by the technical difficulties associated with performing a proper and safe anastomosis with this often smooth organ. Typical postoperative complications, such as pancreatic tistula, leakage of the pancreatic anastomosis, fluid collection and abscess close to this anastomosis, and systemic consequences of these local events, such as sepsis and pulmonary insufficiency, are linked with exocrine pancreatic secretion and the friability of the anastomosis. Although the quality of pancreatic surgery has improved considerably over recent years, mortality fol-
128
163
serum
5-
lipase
4,’ \ ”
_
>\ ‘\
I :I 3
b
%\~*;.- . . . . . . . . . .._._~
0 -1
0
1
2
3
4
5
6
7 days
-
octreotide
-*--placebo
Flgure 1. Serum atnyiase and lipase conwntrations before and after pancreaticresecMn(meanfstandarderrorofthemean)in 125patientsintheoctreotidegroup(sdMHne)and121 patientsin the placebo group (hatched line). The y-axis is manytimes higher
than the normal level for serum. Day 0 represents the serum concentrationin the eveningafter the operation.
THE AMERICAN
JOURNAL
OF SURGERY
VOLUME
JANUARY
1992
OCTREOTIDE TREATMENT FOLLOWING PANCREATIC RESECTION
lowing pancreatic resection is still between 3% and 10% [I -4, and complications continue to occur in 30% to 40% of the patients [1,3,6-181. To overcome the problem of the pancreatic anastomosis, total pancreatectomy [4,Z1,321 or pancreatic duct occlusion [33] were chosen; however, both procedures were for the most part unsuccessful. The concept of inhibiting exocrine secretion perioperatively by means of intravenous somatostatin infusion dates back to 1979. In an open trial, Klempa et al [19] were able to show a drop in postoperative morbidity including the reduced occurrence of acute pancreatitis following Whipple’s resection. Two other groups have used the somatostatin analogue octreotide to prevent pancreatic fistula and acute pancreatitis following pancreatic transplantation [34,35]. It is known that patients suffering from pancreatic and periampullary tumors are at a higher risk of developing postoperative complications than those with chronic pancreatitis [I-7]. This is because in chronic pancreatitis the pancreas usually exhibits a certain amount of tissue fibrosis, which facilitates the anastomosis procedure, in contrast to the often smooth and friable tissue texture in patients harboring tumors in one part of the organ. Therefore, the patients in the present study were stratified into high-risk and low-risk strata according to whether they had tumor disease or a chronically inflamed pancreas. The results confirmed the appropriateness of this stratification in that they showed a considerably higher postoperative morbidity in the patients in the high-risk stratum. We had not expected a statistical difference in mortality because of the number of patients enrolled. However, the trial was designed to determine whether the complication rate, which was estimated to be around 40%, could be reduced by the application of octreotide. Indeed, there was a significant difference in the complication rate between the octreotide group and the placebo group (32% versus 54.4%). This difference was even more pronounced when considering patients in the highrisk stratum. In contrast, when we analyzed patients with chronic pancreatitis (low-risk stratum) separately, the difference in the complication rates did not reach statistical significance in this stratum of 107 patients. This underscores the role of the exocrine pancreatic secretion csipacity, which is reduced in chronic pancreatitis, in the development of postoperative complications following pancreatic resection and the value of the perioperative inhibition of exocrine pancreatic secretion, at least in patients with pancreatic tumors. The most frequent complication following pancreatic resection is pancreatic fistula, which occurred in 10% to 40% of patients retrospectively analyzed for postoperative complications [I-3,5,13,14,18]. This was confirmed in our prospective study in which 27% of patients experienced pancreatic fistula. We were expecting to be able to reduce the occurrence of postoperative pancreatic fistula by means of prophylactic octreotide because somatostatin and its analogues had been shown to be effective in the treatment of pancreatic fistula [36-381. The application THE AMERICAN
of octreotide reduced not only pancreatic fistula resulting directly from uncontrolled postoperative exocrine secretion from the pancreatic remnant, but also other complications, such as abscess, acute pancreatitis, pulmonary insufficiency, shock, and sepsis, which represent local and systemic sequelae of a pancreatic leak. Again, the reduction of these secondary complications was greater in patients with tumors than in those with chronic pancreatitis, which demonstrates the difference in exocrine secretory capacity between the two strata. In conclusion, the perioperative application of the somatostatin analogue octreotide at a daily dosage of 3 X 100 pg subcutaneously in patients undergoing elective pancreatic resection reduces the occurrence of the typical postoperative complications, particularly in patients with pancreatic and periampullary tumors. AFFILIATIONS From the Department of Surgery (MB, HF, HGB), University of Ulm, Ulm, Germany; the Department of Surgery (IK, IB), Zentralkrankenhaus, Bremen, Germany; the Department of Surgery (PH), University of Erlangen, Erlangen, Germany; the Department of Surgery (US), University of Miinster, Miinster, Germany; the Department of Surgery (HB), University of Diisseldorf, Diisseldorf, Germany; the Department of Surgery (AS), University of Giittingen, GGttingen, Germany; the Department of Surgery (DL), University of Greifswald, Greifswald, Germany; the Department of Surgery (RM), Sttidtische Kliniken Kassel, Kassel, Germany; the Department of Surgery (BK), University of Hamburg, Hamburg, Germany; the Department of Surgery (PW), University of Marburg, Marburg, Germany; the Department of Surgery (JW), Zentralklinikum Augsburg, Augsburg, Germany; the Department of Surgery (ELZ), Klinikum Ludwigshafen, Ludwigshafen, Germany; the Department of Surgery (HDS), University of Mannheim, Mannheim, Germany; the Department of Surgery (BR), University of Giepen, Giefien, Germany; the Department of Surgery (PD), Free University of Berlin, Berlin, Germany; the Department of Surgery (KG), University of Innsbruck, Innsbruck, Austria; the Department of Surgery (RT), University of Tiibingen, Tiibingen, Germany; the Department of Surgery (JK), Behring Krankenhaus, Berlin, Germany; the Department of Medical Statistics (WG), University of Ulm, Ulm, Germany; and the Clinical Research Department (HJD, DW), Sandoz AG, Niimberg, Germany. REFERENCES 1. Jordan GL Jr. Pancreaticresectionfor pancreaticcancer. In: HowardJM, Jordan GL Jr., Reber HA, editors. Surgical diseases of the pancreas. Philadelphia: Lea and Febiger, 1987: 666-7 14. 2. Moossa AR, Gadd M, Lavelle-Jones M. Surgical treatment of exocrine pancreatic cancer. In: Go VLW, Gardner JD, Brooks FP, Lebenthal E, DiMagno EP, Scheele GA, editors. The exocrine pancreas. Biology, pathobiology and diseases. New York: Raven Press, 1986: 713-25. 3. Howard JM. Surgical treatment of chronic pancreatitis. In: Howard JM, Jordan GL Jr., Reber HA, editors. Surgical diseases
JOURNAL OF SURGERY
VOLUME 163
JANUARY
1992
129
BijCHLERETAL
of the pancreas. Philadelphia: Lea and Febiger, 1987: 496-521. 4. Rossi RL, Rothschild J, Braasch JW, Munson JL, ReMine SG. Pancreatoduodenectomy in the management of chronic pancreatitis. Arch Surg 1987; 122: 416-20. 5. Crist DW, Sitzmann JV, Cameron JL. Improved hospital morbidity, mortality, and survival after the Whipple procedure. Ann Surg 1987; 206: 358-65. 6. Peters JH, Carey LC. Historical review of pancreaticoduodenectomy. Am J Surg 1991; 161: 219-25. 7. Prinz RA, Greenlee HB. Pancreatic duct drainage in 100 patients with chronic pancreatitis. Ann Surg 1981; 194: 313-20. 8. Neoptolemos JP, Talbot IC, Carr-Locke DL, et al. Treatment and outcome in 52 consecutive cases of ampullary carcinoma. Br J Surg 1987; 74: 957-61. 9. Robertson JFR, Imrie CW, Hole DJ, Carter DC, Blumgart LH. Management of periampullary carcinoma. Br J Surg 1987; 74: 816-9. 10. Moossa AR. Surgical treatment of chronic pancreatitis: an overview. Br J Surg 1987; 74: 661-7. 11. van Heerden JA, McIllrath DC, Ilstrup DM, Weiland LH. Total pancreatectomy for ductal adenocarcinoma of the pancreas: an update. World J Surg 1988; 12: 658-62. 12. Warshaw AL, Swanson RS. What’s new in general surgery. Pancreatic cancer in 1988. Possibilities and probabilities. Ann Surg 1988; 208: 541-53. 13. Trede M, Schwa11 G. The complications of pancreatectomy. Ann Surg 1988; 207: 39-47. 14. Pellegrini CA, Heck CF, Raper S, Way LW. An analysis of the reduced morbidity and mortality rates after pancreaticoduodenectomy. Arch Surg 1989; 124: 778-81. 15. Bittner R, Roscher R, Sati F, Dopfer HP, Scholzel E, Beger HG. Der Einfluj3 von Tumorgro/Ie und Lymphknotenstatus auf die Prognose des Pankreascarcinoms. Chirurg 1989; 60: 240-5. 16. Grace PA, Pitt HA, Longmire WP. Pylorus preserving pancreatoduodenectomy: an overview. Br J Surg 1990; 77: 968-74. 17. Singh SM, Longmire WP, Reber HA. Surgical palliation for pancreatic cancer. Ann Surg 1990; 212: 132-9. 18. Spencer MP, Sarr MG, Nagorney DM. Radical pancreatectomy for pancreaticcancer in theelderly. AnnSurg 1990; 212: 140-3. 19. Klempa J, Schwedes U, Usadel KH. Verhiitung von postoperativen pankreatitischen Komplikationen nach Duodenopankreatektomie durch Somatostatin. Chirurg 1979; 50: 427-32. 20. Bloom SR, Mortimer DH, Thomer MO, et al. Inhibition of gastrin- and gastric acid secretion by growth hormone release inhib iting hormone. Lancet 1974; II: 1106-9. 21. Raptis S, Schlegel W, Lehmann E, Dollinger HC, Zoupas Ch. Effects of somatostatin on the exocrine pancreas and the release of duodenal hormones. Metabolism 1978; 27: 1321-8. 22. Creutzfeldt W, Lankisch PG, Fiilsch UR. Hemmung der Secretin-und Cholecystokinin-Pancreozymin-induzierten Saft- und Enzymsekretion des Pankreas und der Gallenblasenkontraktion bei Menschen durch Somatostatin. Dtsch Med Wochenschr 1975; 100: 1135-8. 23. Kiihler E, Beglinger Ch, Dettwiler S, Whitehouse I, Gyr K. Effect of a new somatostatin analogue on pancreatic function in healthy volunteers. Pancreas 1986; 2: 154-9. 24. Lembcke B, Creutzfeldt W, Schleser S, Ebert R, Shaw C, Koop I. Effect of the somatostatin analogue sandostatin (SMS 201995) on gastrointestinal, pancreatic and biliary function and hormone release in normal men. Digestion 1987; 36: 108-24. 25. Bauer W, Briner U, Doepfner W, et al. SMS 201-995: a very potent and selective octapeptide analogue of somatostatin with pro longed action. Life Sci 1982; 31: 1133-40. 26. Pless J, Bauer W, Briner U, ef al. Chemistry and pharmacology of SMS 201-995, a long-acting octapeptide analogue of somatosta-
130
THE AMERICAN JOURNAL OF SURGERY
tin. Stand J Gastroenterol 1986; 21 (Suppl 119): 55-64. 27. Frie@ H, Bilchler M, Meschenmoser L, et al. Effect of octreotide (SMS 201-995) on exocrine pancreatic function in man [abstract]. Pancreas 1989; 4: 616. 28. Sokal RR, Rolf FJ. Biometry. 2nd ed. San Francisco: WH Freemann, 198 1. 29. Whipple AO, Parsons WB, Mullins CR. Treatment of carcinoma of the ampulla of Vater. Ann Surg 1935; 102: 763-79. 30. Beger HG, Btlchler M, Bittner R, Oettinger W, Roscher R. Duodenum-preserving resection of the head of the pancreas in severe chronic pancreatitis. Ann Surg 1989; 209: 273-8. 31. Taparel D, Susini C, Esteve JP, et al. Somatostatin analogs: Correlation of receptor affinity with inhibition of cyclic AMP formation in pancreatic acinar cells. Peptides 1985; 6: 109-13. 32. Brooks JR. Total pancreatectomy for pancreatic cancer. In: Howard JM, Jordan GL Jr., Reber HA, editors. Surgical diseases of the pancreas. Philadelphia: Lea and Febiger, 1987: 715-24. 33. Gall FP, Gebhardt C. Ein neues Konzept in der Chirurgie der chronischen Pankreatitis. Dtsch Med Wochenschr 1979; 104: 1003-g. 34. Starzl TE, Tcdo S, Tzakis A, et al. Abdominal organ cluster transplantation for the treatment of upper abdominal malignancies. Ann Surg 1989; 210: 374-86. 35. Daloze P, Beauregard H, St. Louis G, et al. Clinical pancreas transplantation: A learning curve of its management. Transplant Proc 1989; 21: 2858-61. 36. Pederzoli P, Bassi C, Falconi M, Albrigo R, Vantini I, Micciolo R. Conservative treatment of external pancreatic fistulas with parenteral nutrition alone or in combination with continuous intravenous infusion of somatostatin, glucagon or calcitonin. Surg Gynecol Obstet 1986; 163: 428-32. 37. Nubiola-Calonge P, Badia JM, Sancho J, Gil MJ, Segura M, Sitges-Serra A. Blind evaluation of the effect of octreotide (SMS 201-995), a somatostatin analogue, on small-bowel tistula output. Lancet 1987; II: 672-4. 38. Prinz RA, Pickleman J, Hoffman JP. Treatment of pancreatic cutaneous fistula with a somatostatin analogue. Am J Surg 1988; 155: 36-42. DISCUSSION D.W. Rattner (Boston, MA): I’m concerned about the mortality rate in the series, which is higher than that
found in current studies from single institutions around the world, and particularly by the fistula rate. One of your comments that, “We cannot improve the surgeon,” does not seem true. It seems to me that these are surgical complications, irrespective of the use of octreotide, and that there is ample evidence from retrospective studies that octreotide does help these fistulae to close more promptly, I wonder if you would comment on the fistula rate and the mortality rate in the series? Aaron S. Fink (Cincinnati, OH): One of the topics you did not address is potential side effects of the therapy itself. This is not a small dose of somatostatin, particularly in those patients who are low risk, i.e., chronic pancreatitis, who may already be endocrine insufficient. Was glucose or insulin therapy a potential problem in these patients, and were there any side effects or consequences? Carlos A. Pellegrini (San Francisco, CA): I am impressed by this study and by the ability of this German group to assemble such a large number of institutions and produce a clinical trial in such a short period of time. I
VOLUME 163 JANUARY 1992
OCTREOTIDE TREATMENT FOLLOWING PANCREATIC RESECTION
would hope that, possibly through the Society for Surgery of the Alimentary Tract, we in America can organize trials of this sort. I congratulate you because I think these sort of data are what we need. I tend to believe all of your data, because they were collected prospectively in a very short period of time. I think some of our retrospective studies don’t reveal as high an incidence of fistulae because fistulae simply aren’t detected on chart reviews. Charles J. Yeo (Baltimore, Md): Somatostatin is known to suppress the growth of certain tumors. You have 139 patients who’ve now been followed for approximately 1 year. Do you have any information on survival in those patients? Does somatostatin provide a benefit in patient survival, as compared to the control group? L. William Traverse (Seattle, WA): In Germany, an anastomosis is not always performed in conjunction with these operations. The pancreatic duct is obliterated with Ethibloc (Ethicon, Norderstedt, Germany) or another compound, and a formal mucosa-to-mucosa anastomosis is not performed. What percent of these patients in the high-risk or low-risk groups did not have a pancreatic anastomosis performed? Perhaps that is the reason for the lower rate of pancreatic leakage. In addition, the size of the duct or the consistency of the gland is important, when a pancreaticojejunostomy is done. Do you have any information on that variable? Keith D. Lillemoe (Baltimore, Md): You have nicely randomized for high-risk and low-risk patients. Do you have any information with respect to specific operations? For example, were the benefits of octreotide with respect to the complication rate improved following the Whipple’s procedure?
THE AMERICAN
Markus Biichler (closing): Dr. Rattner, I don’t think that an overall mortality rate of 4.5% in 245 patients in a multicentered trial is too high, because these patients underwent major pancreatic resections. I reviewed the literature with regard to the incidence of fistulae after major pancreatic resection, and there is very little prospective information. In retrospective trials, the incidence of tistulae is around 10% to 15%. We define a fistula as a concentration of pancreatic enzymes within the drainage fluid of more than three times the concentration of the enzymes in the serum appearing later than 3 days postoperatively. It is a very strict definition. Dr. Fink, we had three patients with side effects in whom we stopped the treatment. Two of them were in the octreotide group, and one was in the placebo group. All suffered from local problems, i.e., itching, etc. Our patients did not experience any major systemic side effects such as glucose imbalance. Dr. Yea, we have no information about long-term survival within this trial, but we have done another study on 22 patients with pancreatic cancer, treating them with octreotide. This treatment has had no impact on survival. Dr. Traverso, 40 of the 246 patients did not undergo pancreatic anastomosis, but they were treated with Ethibloc. The fistula rate in these patients is higher than the rate in the patients receiving an anastomosis. Dr. Lillmoe, I do not have the data to specifically answer your question. However, 152 patients underwent Whipple’s procedure, and, in the group with tumors, 85% underwent Whipple’s procedure. Therefore, I would sup pose that the benefits from octreotide apply to the group undergoing Whipple’s operation.
JOURNAL
OF SURGERY
VOLUME
163 JANUARY
1992
131
Though morbidity and mortality rates following pancreatic resection have improved in recent years, they are still around 35% and 5%, respectively. Typical complications, such as pancreatic fistula, abscess, and subsequent sepsis, are chiefly associated with exocrine pancreatic secretion. In order to clarify whether the perioperative inhibition of exocrine pancreatic secretion prevents complications, we assessed the efficacy of octreotide, a long-acting somatostatin analogue. We conducted a randomized, double-blind, placebo-controlled, multicenter trial in 246 patients undergoing major elective pancreatic surgery. Patients were stratified into a high-risk stratum (limited to patients with pancreatic and periampullary tumors) or low-risk stratum (patients with chronic pancreatitis) . Patients received octreotide (3 X 100 pg) or placebo subcutaneously for 7 days perioperatively. Eleven complications were defined: death, leakage of anastomosis, pancreatic fistula, abscess, fluid collection, shock, sepsis, bleeding, pulmonary insufficiency, renal insufficiency, and postoperative pancreatitis. Two hundred patients underwent pancreatic head resection, 31 patients underwent left resection, and 15 patients had other procedures. The overall mortality rate within 90 days was 4.5%, with 3.2% in the octreotide group and 5.8% in the placebo group. The complication rate was 32% in the patients receiving octreotide (40 of 125 patients) and 55% in patients receiving placebo (67 of 121 patients) (p
A complete list of all affiliations appears on page 129. Requests for reprints should be addressed to Hans G. Beger, MD, FACS, Department of General Surgery, University of Ulm, Steinhoevelstrasse 9, D-7900 Ulm, Germany. Presented at the 32nd Annual Meeting of the Society for Surgery of the Alimentary Tract, New Orleans, Louisiana, May 20-22, 1991.
The perioperative application of octreotide reduces the occurrence of typical postoperative complications after pancreatic resection, particularly in patients with tumors.
hough the mortality rate following pancreatic resecT tion has considerably improved in recent years, it remains between 3% and 10% [I-6]. Complications including medical and surgical morbidity occur in 30% to 40% of the patients [ 1,3,6-181. Major complications after pancreatic resection, some of which can be fatal, such as pancreatic fistula, intra-abdominal fluid collection, ab scess, and subsequent sepsis, are chiefly associated with exocrine pancreatic secretion. The concept of inhibiting exocrine pancreatic secretion to prevent postoperative complications was originated in 1979 [ 191. In an open trial, a continuous intravenous infusion (250 pg/h) of somatostatin was administered to patients perioperatively after Whipple’s resection. The authors of the study reported a reduced complication rate. The human hormone somatostatin has been demonstrated to strongly inhibit basal and stimulated exocrine pancreatic secretion [20-221. Recently, the synthetic somatostatin analogue, octreotide (SMS 201995), has become clinically available, and it is more potent in inhibiting pancreatic enzyme secretion [23,24] and has a much longer half-life [25,26] than does the native hormone. In a multicenter controlled and randomized trial, we analyzed the value of the perioperative application (7 days) of octreotide in patients undergoing major elective pancreatic surgery. PATIENTS
AND METHODS
Between October 1989 and July 1990, 322 patients were recruited from 18 surgical departments in Germany (n = 17) and Austria (n = 1). The study was approved by the ethics committees of each of the participating hospitals. Patients with pancreatic or periampullary tumors and those suffering from chronic pancreatitis who were suitable for elective pancreatic surgery (resection) were enrolled in the study after giving written, informed consent. Patients were excluded who were undergoing emergency surgery of the pancreas (acute pancreatitis, pancreatic trauma), total pancreatectomy, or pancreatic transplantation, or who were receiving only an elective pancreatic-cyst anastomosis. On the day before the operation, the patients who satisfied the inclusion criteria were randomly assigned to receive a 7-day treatment of
THE AMERICAN JOURNAL OF SURGERY
VOLUME 163 JANUARY 1992
125
BtiCHLER
ET AL
either 3 X 100 pg octreotide (1 mL volume) subcutaneously (every 8 hours) or 21 subcutaneous placebo injections (1 mL each) in a double-blinded fashion. Treatment was initiated at 7:00 AM on the day of operation, in each patient at least 1 hour before the laparotomy. The dosage of 3 X 100 pg subcutaneously per day was chosen after a pilot trial [27] had shown that this dosage effects maximum inhibition of pancreatic enzyme secretion in humans. The patients were stratified into two groups: those with pancreatic or periampullary tumors (high-risk stratum) and those suffering from chronic pancreatitis (low-risk stratum). Randomization was carried out separately for each hospital and for each stratum. Each patient received perioperative intravenous antibiotic prophylaxis for 24 to 48 hours using bactericidal drugs effective against gram-negative bacteria. The patients were treated in an intensive care unit for at least 3 days postoperatively. Intravenous crystalloid solutions were administered (2.5 to 3.0 L/day) according to clinical requirements and were supplemented by plasma or blood components in order to maintain a positive central venous pressure and a blood hematocrit greater than or equal to 30%. The patients were permitted nothing orally for at least 3 postoperative days. Definition of postoperative complications: The study protocol defined 11 typical complications occurring after pancreatic resection and stipulated that each patient be followed for 90 days postoperatively. The complications were defined as follows: (1) Death (within 90 days postoperatively). (2) Leakage of pancreatic, biliary, or intestinal anastomosis as determined by radiograph or intraoperative (relaparotomy) findings. (3) Pancreatic fistula was defined by a concentration of amylase and lipase in the drainage fluid more than 3 days postoperatively of more than three times the serum concentration and a drainage volume of more than 10 mL/24 hours at the same time. The serum and drainage fluid amylase and lipase concentration was determined on postoperative days 1, 3,4, 5, and 7 and twice a week thereafter. (4) An intra-abdominal abscess was defined as a pus collection or an infected fluid collection revealed by ultrasound or computed tomography (CT) guided needle aspiration and microbiological culture or demonstrated by relaparotomy. (5) An intra-abdominal fluid collection (sterile) was defined as a collection of fluid of at least 5 X 5 cm in diameter in ultrasound or CT that did not qualify as an abscess. All patients were screened for a fluid collection by means of ultrasound or CT on day 4 and day 8 postoperatively. (6) Cardiocirculatory shock was defined as a systolic blood pressure of less than 80 mm Hg for more than 20 minutes (later than 24 hours postoperatively) that required continuous volume replacement and/or treatment with catecholamines. (7) Sepsis was defined as the presence of at least four of the following five criteria: (A) bacterial-positive blood culture, (B) rectal temperature greater than 38.5% (C)
leukocytosis greater than 12,000/mm3 or leukopenia less than 3,000/mm3, (D) thrombocytopenia less than 100,000/mm3, and (E) metabolic acidosis (base excess greater than -4). Blood cultures were taken daily in patients with rectal temperature of higher than 38.5’C. Blood leukocyte and thrombocyte determination and a blood gas analysis were performed on postoperative days 1, 3, 4, 5, and 7 and twice a week thereafter. (8) Pulmonary insufficiency was defined as an arterial PO2 of less than 60 mm Hg, despite oxygen application (4 L/m) via the nose, or when the patient required prolonged postoperative mechanical ventilation (greater than 24 hours after the end of surgery) or reintubation. (9) Renal insufficiency was defined as an increase in serum creatinine of more than twice the preoperative value. Patients whose serum creatinine was increased preoperatively (greater than 120 pmol/L) were not included in the analysis of this criterion. Serum creatinine was monitored on postoperative days 1, 3, 4, 5, and 7 and twice a week thereafter. (10) Bleeding was defined as the replacement of more than 3 units of blood more than 24 hours after the end of surgery or the indication for relaparotomy due to intraabdominal or intra-intestinal bleeding. Blood hematocrit was monitored daily on postoperative days 1,3,4,5, and 7 and twice a week thereafter. (11) Postoperative acute pancreatitis was defined as a more than threefold increase in serum amylase or lipase more than 3 days after surgery, accompanied by morphological signs of acute pancreatitis in a contrast-enhanced CT scan. CT was mandatory in patients with a threefold increase in serum amylase or lipase on postoperative day 4 or after. Statistical analysis: During the design of the study, we anticipated a rate of 30% to 40% patients with one or more complications in the placebo group. A power calculation for a two-tailed significance level of 0.05 indicated that it would be reasonable to recruit about 100 patients for each group. This number of patients enabled the study to detect a reduction of complication rate from 30% to 40% in the placebo group to 15% to 20% in the octreotide group with a power of 90%. The hypothesis testing was done with the Wilcoxon two-sample rank sum test for qualitative outcome variables, with the chi-square test for quantitative outcome variables. For testing small subgroups, the chi-square test was corrected for discontinuity as proposed by Williams [28]. The hypotheses for the comparison of the complication rates between the placebo and octreotide groups were formulated a priori, making this test strictly confirmatory.
126
163
THE AMERICAN
JOURNAL
OF SURGERY
VOLUME
RESULTS
On the day before the operation, 322 patients were randomized and received either a primary subcutaneous injection of placebo or 100 pg of octreotide in the morning of the day of surgery. During the operation, 246 of the 322 patients (76.4%) were found to be operable (resectable), according to the conditions stipulated in the study protocol. Seventy-six subjects were withdrawn from the study, and the injection therapy was discontinued beJANUARY
1992
DCTREOTIDE TREATMENT FOLLOWING PANCREATIC RESECTION
TABLE
TABLE
I
II Mortality and Complications
Comparablllty of Groups
Age (median, range) Sex (m/f) Type of disease Pancreatic cancer Periampullary cancer Endocrine tumor Chronic pancreatitis Others
Octreotide (n = 125)
Placebo (n = 121)
51 (19-75) 92133
52 (20-80) 82139
37 15 5 82 8
Complication rate (patients with 1 or more complications) Number of complications (Total 208)
34 25 4 50 8
73 26
79 22
18 3 2 3
13 5 1 1
4 (3.2%)
7 (5.8%)
40 (32.0%) p < 0.005 67 (55.4%) 126
82
III
Number and Type of Complications Octreotide (n = 125)
cause pancreatic resection was not possible due to advanced tumor disease (n = 73) or because only a pancreatic-cyst anastomosis was carried out (n = 3). In three patients (two from the octreotide group, one from the placebo group), therapy was discontinued after 2 and 4 days, respectively, due to evidence of side effects. The three patients were retained in the trial. A total of 246 patients were included in the final evaluation, with 125 subjects assigned to the octreotide group and 12 1 to the placebo group. The two groups were well matched with regard to age, sex, type of underlying disease, and type of surgery performed (Table I). F’reoperative morbidity (heart, lung, liver, kidney, diabetes mellitus) were comparable between the two groups. Of the 246 patients, 139 were in the high-risk stratum (tumors) and 107 in the low-risk stratum (chronic pancreatitis). Pancreaticoduodenectomy (Whipple’s operation [29]) was performed in 152 patients, a duodenumpreserving pancreatic head resection [30] in 48 patients, a pancreatic left resection in 3 1 patients, pancreaticojejunostomy in 8 patients, enucleation for endocrine tumor in 3 patients, and atypical pancreatic resection in 4 patients (Table I). The overall mortality within 90 days of surgery was 4.5% (11 of 246 patients). -Four patients (3.2%) of the octreotide group and 7 patients (5.8%) of the placebo group died (Table II). Among the high-risk patients (139 patients), the mortality rates were 2.9% and 9.9% in octreotide and placebo groups, respectively. Forty patients (32%) of the octreotide group and 67 (55.4%) of the placebo group suffered one or more postoperative complications as defined in the study protocol (Table II). This difference was significant at a level of p <0.005. The total complication rate was 43.5% (107 of 246 patients). In these 107 patients, a total of 208 complications were recorded, 82 and 126 in the octreotide and placebo groups, respectively (Table II). The most frequent complication was pancreatic fwtula (27%), followed by pulTHE AMERICAN
Placebo (n = 121)
Mortality
TABLE Type of operation Whipple’s Duodenum-preserving pancreatic head resection Left resection Pancreaticojejunostomy Enucleation Others
Octreotide (n = 125)
4 6 22 8 8 6 3 12 1 12 0 82
Death Leakage of anastomosis Pancreatic fistula Abscess Fluid collection Shock Sepsis Pulmonary insufficiency Renal insufficiency Bleeding Postoperative pancreatiiis Total complications
TABLE
Placebo (n = 121) 7 7 46 12 9 7 6 16 2 10 4 126
IV
Number of Patients With One or More Complications in High-Risk and Low-Risk Strata Octreotide
Placebo
No. of high-risk* patients (n = 139)
68
71
No. of high-risk patients with one or more complications
26 (38%)
No. of low-riskf patients (n = 107)
57
No. of low-risk patients with one or more complications
14 (25%)
‘Patients with pancreatic or periampullary tPatients with chronic pancreatitis.
p co.01
46 (65%) 50
NS
21 (42%)
tumors.
monary insufficiency (1 l%), bleeding (9%), abscess (8%), fluid collection (7%), anastomosis leakage (5%) shock (5%), death (4.5%), sepsis (4%), pancreatitis (2%), and renal insufficiency ( 1%). The frequency of death, pancreatic listula, abscess, sepsis, pulmonary insufficiency, and postoperative pancreatitis was lower in the patients in octreotide group than in the placebo group (Table III). The rates of complications among the high-risk patients were 38% for the octreotide group and 65% for the placebo group (p
JOURNAL
OF SURGERY
VOLUME
163
JANUARY
1992
127
BikHLER
TABLE
ET AL
V
Number and Type of Complications in Patients in the High-Risk Stratum (Tumors) Octreotide (n = 66)
Placebo (n = 71)
2 4 16 4 5 4 1 7 0 7 0
7 6 29 IO 7 7 5 12 2 9 4 96
Death Leakage of anastomosis Pancreatic fistula Abscess Fluid collection Shock Sepsis Pulmonary insufficiency Renal insufficiency Bleeding Postoperative pancreatitis Total number of complications
4
after the operation) and 1 (lipase and amylase). This was lower in the patients receiving octreotide than in those in the placebo group (Figure 1). The mean hospitalization time of the 246 evaluable patients was 24.1 f 1.2 days. The patients in the octreotide and placebo groups were hospitalized for 22.1 f 1.5 and 26.2 f 1.9 days, respectively. In patients in the highrisk stratum who had a mean hospitalization of 32.2 f 2.6 days, those in the octreotide group remained in the hospital 29.0 f 2.4 days and in the placebo group 35.3 f 2.9 days. Fifty-six patients (31 octreotide, 25 placebo) experienced pain during the subcutaneous injection at the injection site, which did not require discontinuation of the treatment. Two patients (both in the octreotide group) suffered hot flashes, and the treatment was stopped on day 2 and 4, respectively. One patient (in the placebo group) had an inflammatory reaction at the injection site that led to discontinuation of the treatment on day 4. Two other patients suffered from itching and temporary migrant exanthema, which did not necessitate discontinuation of the treatment. No further side effects were registered.
evidenced by the number and type of complications (Table V). Table V also shows a difference in the occurrence of death, pancreatic fistula, abscess, shock, sepsis, pulmonary insufficiency, and postoperative pancreatitis. Serum pancreatic enzyme values showed a significant difference on postoperative days 0 (lipase) (in the evening
COMMENTS Somatostatin, a tetradecapeptide originally discovered in hypothalamic tissue, is a powerful inhibitor of baseline and stimulated exocrine pancreatic secretion [20-221. Its action on the pancreas is mediated via the inhibition of secretagogue gastrointestinal hormones and by direct pathways via somatostatin receptors at acinus cells [31]. Octreotide (SMS 201-995), an octapeptide analogue of somatostatin, was synthesized to have morespecific, more-potent, and longer-acting inhibitory effects than native somatostatin [25,26j. The longer half-life of octreotide (100 min), in particular, in comparison with somatostatin 14 (2 min) has made this compound advantageous for clinical application. In a pilot trial, we were able to show that 100 pg octreotide administered subcutaneously inhibited the stimulated secretion of amylase, trypsin, and chymotrypsin by 84%,76%, and 77%, respectively [27]. Our data confirm those of earlier reports that showed a maximum inhibition of pancreatic protease and amylase secretion with comparable dosages of octreotide [23,24]. In pancreatic surgery, the pancreatic remnant and its anastomosis with stomach or small bowel constitutes the major risk factor. This problem is caused by a continuous exocrine pancreatic secretion (including the secretion of harmful proteases and lipolytic enzymes) of the pancreatic stump and by the technical difficulties associated with performing a proper and safe anastomosis with this often smooth organ. Typical postoperative complications, such as pancreatic tistula, leakage of the pancreatic anastomosis, fluid collection and abscess close to this anastomosis, and systemic consequences of these local events, such as sepsis and pulmonary insufficiency, are linked with exocrine pancreatic secretion and the friability of the anastomosis. Although the quality of pancreatic surgery has improved considerably over recent years, mortality fol-
128
163
serum
5-
lipase
4,’ \ ”
_
>\ ‘\
I :I 3
b
%\~*;.- . . . . . . . . . .._._~
0 -1
0
1
2
3
4
5
6
7 days
-
octreotide
-*--placebo
Flgure 1. Serum atnyiase and lipase conwntrations before and after pancreaticresecMn(meanfstandarderrorofthemean)in 125patientsintheoctreotidegroup(sdMHne)and121 patientsin the placebo group (hatched line). The y-axis is manytimes higher
than the normal level for serum. Day 0 represents the serum concentrationin the eveningafter the operation.
THE AMERICAN
JOURNAL
OF SURGERY
VOLUME
JANUARY
1992
OCTREOTIDE TREATMENT FOLLOWING PANCREATIC RESECTION
lowing pancreatic resection is still between 3% and 10% [I -4, and complications continue to occur in 30% to 40% of the patients [1,3,6-181. To overcome the problem of the pancreatic anastomosis, total pancreatectomy [4,Z1,321 or pancreatic duct occlusion [33] were chosen; however, both procedures were for the most part unsuccessful. The concept of inhibiting exocrine secretion perioperatively by means of intravenous somatostatin infusion dates back to 1979. In an open trial, Klempa et al [19] were able to show a drop in postoperative morbidity including the reduced occurrence of acute pancreatitis following Whipple’s resection. Two other groups have used the somatostatin analogue octreotide to prevent pancreatic fistula and acute pancreatitis following pancreatic transplantation [34,35]. It is known that patients suffering from pancreatic and periampullary tumors are at a higher risk of developing postoperative complications than those with chronic pancreatitis [I-7]. This is because in chronic pancreatitis the pancreas usually exhibits a certain amount of tissue fibrosis, which facilitates the anastomosis procedure, in contrast to the often smooth and friable tissue texture in patients harboring tumors in one part of the organ. Therefore, the patients in the present study were stratified into high-risk and low-risk strata according to whether they had tumor disease or a chronically inflamed pancreas. The results confirmed the appropriateness of this stratification in that they showed a considerably higher postoperative morbidity in the patients in the high-risk stratum. We had not expected a statistical difference in mortality because of the number of patients enrolled. However, the trial was designed to determine whether the complication rate, which was estimated to be around 40%, could be reduced by the application of octreotide. Indeed, there was a significant difference in the complication rate between the octreotide group and the placebo group (32% versus 54.4%). This difference was even more pronounced when considering patients in the highrisk stratum. In contrast, when we analyzed patients with chronic pancreatitis (low-risk stratum) separately, the difference in the complication rates did not reach statistical significance in this stratum of 107 patients. This underscores the role of the exocrine pancreatic secretion csipacity, which is reduced in chronic pancreatitis, in the development of postoperative complications following pancreatic resection and the value of the perioperative inhibition of exocrine pancreatic secretion, at least in patients with pancreatic tumors. The most frequent complication following pancreatic resection is pancreatic fistula, which occurred in 10% to 40% of patients retrospectively analyzed for postoperative complications [I-3,5,13,14,18]. This was confirmed in our prospective study in which 27% of patients experienced pancreatic fistula. We were expecting to be able to reduce the occurrence of postoperative pancreatic fistula by means of prophylactic octreotide because somatostatin and its analogues had been shown to be effective in the treatment of pancreatic fistula [36-381. The application THE AMERICAN
of octreotide reduced not only pancreatic fistula resulting directly from uncontrolled postoperative exocrine secretion from the pancreatic remnant, but also other complications, such as abscess, acute pancreatitis, pulmonary insufficiency, shock, and sepsis, which represent local and systemic sequelae of a pancreatic leak. Again, the reduction of these secondary complications was greater in patients with tumors than in those with chronic pancreatitis, which demonstrates the difference in exocrine secretory capacity between the two strata. In conclusion, the perioperative application of the somatostatin analogue octreotide at a daily dosage of 3 X 100 pg subcutaneously in patients undergoing elective pancreatic resection reduces the occurrence of the typical postoperative complications, particularly in patients with pancreatic and periampullary tumors. AFFILIATIONS From the Department of Surgery (MB, HF, HGB), University of Ulm, Ulm, Germany; the Department of Surgery (IK, IB), Zentralkrankenhaus, Bremen, Germany; the Department of Surgery (PH), University of Erlangen, Erlangen, Germany; the Department of Surgery (US), University of Miinster, Miinster, Germany; the Department of Surgery (HB), University of Diisseldorf, Diisseldorf, Germany; the Department of Surgery (AS), University of Giittingen, GGttingen, Germany; the Department of Surgery (DL), University of Greifswald, Greifswald, Germany; the Department of Surgery (RM), Sttidtische Kliniken Kassel, Kassel, Germany; the Department of Surgery (BK), University of Hamburg, Hamburg, Germany; the Department of Surgery (PW), University of Marburg, Marburg, Germany; the Department of Surgery (JW), Zentralklinikum Augsburg, Augsburg, Germany; the Department of Surgery (ELZ), Klinikum Ludwigshafen, Ludwigshafen, Germany; the Department of Surgery (HDS), University of Mannheim, Mannheim, Germany; the Department of Surgery (BR), University of Giepen, Giefien, Germany; the Department of Surgery (PD), Free University of Berlin, Berlin, Germany; the Department of Surgery (KG), University of Innsbruck, Innsbruck, Austria; the Department of Surgery (RT), University of Tiibingen, Tiibingen, Germany; the Department of Surgery (JK), Behring Krankenhaus, Berlin, Germany; the Department of Medical Statistics (WG), University of Ulm, Ulm, Germany; and the Clinical Research Department (HJD, DW), Sandoz AG, Niimberg, Germany. REFERENCES 1. Jordan GL Jr. Pancreaticresectionfor pancreaticcancer. In: HowardJM, Jordan GL Jr., Reber HA, editors. Surgical diseases of the pancreas. Philadelphia: Lea and Febiger, 1987: 666-7 14. 2. Moossa AR, Gadd M, Lavelle-Jones M. Surgical treatment of exocrine pancreatic cancer. In: Go VLW, Gardner JD, Brooks FP, Lebenthal E, DiMagno EP, Scheele GA, editors. The exocrine pancreas. Biology, pathobiology and diseases. New York: Raven Press, 1986: 713-25. 3. Howard JM. Surgical treatment of chronic pancreatitis. In: Howard JM, Jordan GL Jr., Reber HA, editors. Surgical diseases
JOURNAL OF SURGERY
VOLUME 163
JANUARY
1992
129
BijCHLERETAL
of the pancreas. Philadelphia: Lea and Febiger, 1987: 496-521. 4. Rossi RL, Rothschild J, Braasch JW, Munson JL, ReMine SG. Pancreatoduodenectomy in the management of chronic pancreatitis. Arch Surg 1987; 122: 416-20. 5. Crist DW, Sitzmann JV, Cameron JL. Improved hospital morbidity, mortality, and survival after the Whipple procedure. Ann Surg 1987; 206: 358-65. 6. Peters JH, Carey LC. Historical review of pancreaticoduodenectomy. Am J Surg 1991; 161: 219-25. 7. Prinz RA, Greenlee HB. Pancreatic duct drainage in 100 patients with chronic pancreatitis. Ann Surg 1981; 194: 313-20. 8. Neoptolemos JP, Talbot IC, Carr-Locke DL, et al. Treatment and outcome in 52 consecutive cases of ampullary carcinoma. Br J Surg 1987; 74: 957-61. 9. Robertson JFR, Imrie CW, Hole DJ, Carter DC, Blumgart LH. Management of periampullary carcinoma. Br J Surg 1987; 74: 816-9. 10. Moossa AR. Surgical treatment of chronic pancreatitis: an overview. Br J Surg 1987; 74: 661-7. 11. van Heerden JA, McIllrath DC, Ilstrup DM, Weiland LH. Total pancreatectomy for ductal adenocarcinoma of the pancreas: an update. World J Surg 1988; 12: 658-62. 12. Warshaw AL, Swanson RS. What’s new in general surgery. Pancreatic cancer in 1988. Possibilities and probabilities. Ann Surg 1988; 208: 541-53. 13. Trede M, Schwa11 G. The complications of pancreatectomy. Ann Surg 1988; 207: 39-47. 14. Pellegrini CA, Heck CF, Raper S, Way LW. An analysis of the reduced morbidity and mortality rates after pancreaticoduodenectomy. Arch Surg 1989; 124: 778-81. 15. Bittner R, Roscher R, Sati F, Dopfer HP, Scholzel E, Beger HG. Der Einfluj3 von Tumorgro/Ie und Lymphknotenstatus auf die Prognose des Pankreascarcinoms. Chirurg 1989; 60: 240-5. 16. Grace PA, Pitt HA, Longmire WP. Pylorus preserving pancreatoduodenectomy: an overview. Br J Surg 1990; 77: 968-74. 17. Singh SM, Longmire WP, Reber HA. Surgical palliation for pancreatic cancer. Ann Surg 1990; 212: 132-9. 18. Spencer MP, Sarr MG, Nagorney DM. Radical pancreatectomy for pancreaticcancer in theelderly. AnnSurg 1990; 212: 140-3. 19. Klempa J, Schwedes U, Usadel KH. Verhiitung von postoperativen pankreatitischen Komplikationen nach Duodenopankreatektomie durch Somatostatin. Chirurg 1979; 50: 427-32. 20. Bloom SR, Mortimer DH, Thomer MO, et al. Inhibition of gastrin- and gastric acid secretion by growth hormone release inhib iting hormone. Lancet 1974; II: 1106-9. 21. Raptis S, Schlegel W, Lehmann E, Dollinger HC, Zoupas Ch. Effects of somatostatin on the exocrine pancreas and the release of duodenal hormones. Metabolism 1978; 27: 1321-8. 22. Creutzfeldt W, Lankisch PG, Fiilsch UR. Hemmung der Secretin-und Cholecystokinin-Pancreozymin-induzierten Saft- und Enzymsekretion des Pankreas und der Gallenblasenkontraktion bei Menschen durch Somatostatin. Dtsch Med Wochenschr 1975; 100: 1135-8. 23. Kiihler E, Beglinger Ch, Dettwiler S, Whitehouse I, Gyr K. Effect of a new somatostatin analogue on pancreatic function in healthy volunteers. Pancreas 1986; 2: 154-9. 24. Lembcke B, Creutzfeldt W, Schleser S, Ebert R, Shaw C, Koop I. Effect of the somatostatin analogue sandostatin (SMS 201995) on gastrointestinal, pancreatic and biliary function and hormone release in normal men. Digestion 1987; 36: 108-24. 25. Bauer W, Briner U, Doepfner W, et al. SMS 201-995: a very potent and selective octapeptide analogue of somatostatin with pro longed action. Life Sci 1982; 31: 1133-40. 26. Pless J, Bauer W, Briner U, ef al. Chemistry and pharmacology of SMS 201-995, a long-acting octapeptide analogue of somatosta-
130
THE AMERICAN JOURNAL OF SURGERY
tin. Stand J Gastroenterol 1986; 21 (Suppl 119): 55-64. 27. Frie@ H, Bilchler M, Meschenmoser L, et al. Effect of octreotide (SMS 201-995) on exocrine pancreatic function in man [abstract]. Pancreas 1989; 4: 616. 28. Sokal RR, Rolf FJ. Biometry. 2nd ed. San Francisco: WH Freemann, 198 1. 29. Whipple AO, Parsons WB, Mullins CR. Treatment of carcinoma of the ampulla of Vater. Ann Surg 1935; 102: 763-79. 30. Beger HG, Btlchler M, Bittner R, Oettinger W, Roscher R. Duodenum-preserving resection of the head of the pancreas in severe chronic pancreatitis. Ann Surg 1989; 209: 273-8. 31. Taparel D, Susini C, Esteve JP, et al. Somatostatin analogs: Correlation of receptor affinity with inhibition of cyclic AMP formation in pancreatic acinar cells. Peptides 1985; 6: 109-13. 32. Brooks JR. Total pancreatectomy for pancreatic cancer. In: Howard JM, Jordan GL Jr., Reber HA, editors. Surgical diseases of the pancreas. Philadelphia: Lea and Febiger, 1987: 715-24. 33. Gall FP, Gebhardt C. Ein neues Konzept in der Chirurgie der chronischen Pankreatitis. Dtsch Med Wochenschr 1979; 104: 1003-g. 34. Starzl TE, Tcdo S, Tzakis A, et al. Abdominal organ cluster transplantation for the treatment of upper abdominal malignancies. Ann Surg 1989; 210: 374-86. 35. Daloze P, Beauregard H, St. Louis G, et al. Clinical pancreas transplantation: A learning curve of its management. Transplant Proc 1989; 21: 2858-61. 36. Pederzoli P, Bassi C, Falconi M, Albrigo R, Vantini I, Micciolo R. Conservative treatment of external pancreatic fistulas with parenteral nutrition alone or in combination with continuous intravenous infusion of somatostatin, glucagon or calcitonin. Surg Gynecol Obstet 1986; 163: 428-32. 37. Nubiola-Calonge P, Badia JM, Sancho J, Gil MJ, Segura M, Sitges-Serra A. Blind evaluation of the effect of octreotide (SMS 201-995), a somatostatin analogue, on small-bowel tistula output. Lancet 1987; II: 672-4. 38. Prinz RA, Pickleman J, Hoffman JP. Treatment of pancreatic cutaneous fistula with a somatostatin analogue. Am J Surg 1988; 155: 36-42. DISCUSSION D.W. Rattner (Boston, MA): I’m concerned about the mortality rate in the series, which is higher than that
found in current studies from single institutions around the world, and particularly by the fistula rate. One of your comments that, “We cannot improve the surgeon,” does not seem true. It seems to me that these are surgical complications, irrespective of the use of octreotide, and that there is ample evidence from retrospective studies that octreotide does help these fistulae to close more promptly, I wonder if you would comment on the fistula rate and the mortality rate in the series? Aaron S. Fink (Cincinnati, OH): One of the topics you did not address is potential side effects of the therapy itself. This is not a small dose of somatostatin, particularly in those patients who are low risk, i.e., chronic pancreatitis, who may already be endocrine insufficient. Was glucose or insulin therapy a potential problem in these patients, and were there any side effects or consequences? Carlos A. Pellegrini (San Francisco, CA): I am impressed by this study and by the ability of this German group to assemble such a large number of institutions and produce a clinical trial in such a short period of time. I
VOLUME 163 JANUARY 1992
OCTREOTIDE TREATMENT FOLLOWING PANCREATIC RESECTION
would hope that, possibly through the Society for Surgery of the Alimentary Tract, we in America can organize trials of this sort. I congratulate you because I think these sort of data are what we need. I tend to believe all of your data, because they were collected prospectively in a very short period of time. I think some of our retrospective studies don’t reveal as high an incidence of fistulae because fistulae simply aren’t detected on chart reviews. Charles J. Yeo (Baltimore, Md): Somatostatin is known to suppress the growth of certain tumors. You have 139 patients who’ve now been followed for approximately 1 year. Do you have any information on survival in those patients? Does somatostatin provide a benefit in patient survival, as compared to the control group? L. William Traverse (Seattle, WA): In Germany, an anastomosis is not always performed in conjunction with these operations. The pancreatic duct is obliterated with Ethibloc (Ethicon, Norderstedt, Germany) or another compound, and a formal mucosa-to-mucosa anastomosis is not performed. What percent of these patients in the high-risk or low-risk groups did not have a pancreatic anastomosis performed? Perhaps that is the reason for the lower rate of pancreatic leakage. In addition, the size of the duct or the consistency of the gland is important, when a pancreaticojejunostomy is done. Do you have any information on that variable? Keith D. Lillemoe (Baltimore, Md): You have nicely randomized for high-risk and low-risk patients. Do you have any information with respect to specific operations? For example, were the benefits of octreotide with respect to the complication rate improved following the Whipple’s procedure?
THE AMERICAN
Markus Biichler (closing): Dr. Rattner, I don’t think that an overall mortality rate of 4.5% in 245 patients in a multicentered trial is too high, because these patients underwent major pancreatic resections. I reviewed the literature with regard to the incidence of fistulae after major pancreatic resection, and there is very little prospective information. In retrospective trials, the incidence of tistulae is around 10% to 15%. We define a fistula as a concentration of pancreatic enzymes within the drainage fluid of more than three times the concentration of the enzymes in the serum appearing later than 3 days postoperatively. It is a very strict definition. Dr. Fink, we had three patients with side effects in whom we stopped the treatment. Two of them were in the octreotide group, and one was in the placebo group. All suffered from local problems, i.e., itching, etc. Our patients did not experience any major systemic side effects such as glucose imbalance. Dr. Yea, we have no information about long-term survival within this trial, but we have done another study on 22 patients with pancreatic cancer, treating them with octreotide. This treatment has had no impact on survival. Dr. Traverso, 40 of the 246 patients did not undergo pancreatic anastomosis, but they were treated with Ethibloc. The fistula rate in these patients is higher than the rate in the patients receiving an anastomosis. Dr. Lillmoe, I do not have the data to specifically answer your question. However, 152 patients underwent Whipple’s procedure, and, in the group with tumors, 85% underwent Whipple’s procedure. Therefore, I would sup pose that the benefits from octreotide apply to the group undergoing Whipple’s operation.
JOURNAL
OF SURGERY
VOLUME
163 JANUARY
1992
131