Pylorus preserving pancreatoduodenectomy: An overview

zyx z zyxwvut zyxwvutsrqpo Review Br. J. Surg. 1990, Vol. 77, September, 968-974 PyIo rus prese rving pancreatoduodenectomy: an overview Pylorus preserving pancreatoduodenectomy (PPPD) was reintroduced I2 years ago. Since that time, over 400 patients have undergone PPPD with approximately 41 per cent having chronic pancreatitis and 54 per cent having pancreatic and other periampullary malignancies. Reported 5-year survivals in this latter group have been comparable to those achieved by the classic Whipple procedure. The postoperative mortality rate in 339 reported patients has been 3-8 per cent. Postoperative morbidity, including delayed gastric emptying, has been similar to that of the classic Whipple operation. However, PPPD has been associated with fewer late problems with dumping, diarrhoea, delayed gastric emptying (8.6 per cent), and marginal ulceration (3.6 per cent). Moreover, most patients undergoing PPPD have been able to return to their preoperative andpreillness weight. The additional advantage of decreased operative time makes PPPD an attractive alternative to the classic pancreatoduodenectomy. zyxwvutsrqpo zyxwvutsrqpo zyxwv zyxwvuts zyxwvutsrqpo P. A. Grace*, H. A. P i t t t and W. P. Longmire1 *Department of Surgery, Royal College of Surgeons in Ireland, Dublin, Ireland and ?Department of Surgery, The Johns Hopkins Medical Institutions, Baltimore, Maryland, USA and $Department of Surgery, UCLA Medical Center, Los Angeles, California, USA Correspondence to: Mr H. A. Pitt, The Johns Hopkins Hospital, 600 North Wolfe Street, Blalock 688, Baltimore, Maryland 21205, USA Keywords: Pancreatoduodenectomy, chronic pancreatitis, pancreatic carcinoma, arnpullary carcinoma, distal bile duct carcinoma, duodenal carcinoma The incidence of pancreatic cancer and periampullary tumours has increased steadily over the last 40 years’ while chronic pancreatitis, although an uncommon condition*, continues to pose a major therapeutic challenge to the surgeon3. In recent years a variety of new surgical approaches to pancreatic resection have been developed. These approaches range from the radical regional pancreatoduodenectomy advocated by Fortner4 to the pylorus preserving pancreatoduodenectomy ( P P P D ) and duodenal preserving pancreatectomy popularized by Longmire’ and Russell6, respectively. Local excision of periampullary carcinomas has also had its advocates in recent years’. The purpose of the present analysis is to review the impact of PPPD on pancreatic surgery since its reintroduction 12 years ago. Historical development While there are several reports of pancreatoduodenal resections from the end of the 19th and the beginning of the 20th centuries8-”, it was not until 1935 that the modern era of pancreatic surgery was inaugurated. In that year Whipple, Parsons and M ~ l l e n s ’published ~ their report of the first successful two-stage radical pancreatoduodenectomy for periampullary carcinoma. In the first stage of the operation, the patient underwent a gastrojejunostomy, ligation of the distal common bile duct and cholecystogastrostomy. Subsequently, the second part of the duodenum, the ampulla of Vater, the adjacent head of the pancreas and the distal common bile duct were removed, and the pancreatic stump was oversewn. Thus, the antrum and pylorus were structurally preserved as a blind stump, and the patency of the gastrointestinal tract was restored by a gastroenterostomy. advocated resection of the distal In 1941 Trimble et stomach with a Polya antecolic anastomosis, eliminating the blind pyloric stump and the suture line in the duodenum, a potential source of leakage. Whipple also advocated distal gastrectomy and a choledochojejunostomy, and both Whipple and Trimble independently reported successful one stage radical 968 pancreatoduodenectomy in 1941’4*’5. I n 1944 Watson16 described a pancreatoduodenectomy for carcinoma of the ampulla of Vater in which he preserved the antrum, pylorus, and 1 inch of duodenum. H e re-established alimentary continuity with a duodenojejunostomy thus preserving in functional continuity the antrum, pylorus, and first part of the duodenum. Watson believed that preservation of the stomach would lead to better digestion and that a duodenojejunostomy would prevent anastomotic ulceration. To Watson, therefore, must go the credit for describing the first PPPD. Watson’s report, however, was forgotten, and gastric resection with or without vagotomy became an integral part of the standard The rationale for inclusion of ‘Whipple’ operation’ ’-’O. gastrectomy and vagotomy has been to prevent marginal ulceration and to provide adequate margins when resecting tumours. This operation continues to be the most frequently performed operation for resectable lesions in the head of the pancreas. In 1978 Traverso and Longmire5 revived the idea of preserving the pylorus during pancreatoduodenectomy. They reported two patients in whom the stomach, pylorus, and first part of the duodenum were preserved and alimentary tract continuity was re-established by end-to-side duodenojejunostomy. These authors, like Watson, reasoned that preservation of the stomach and pylorus would eliminate the complications of a reduced gastric reservoir and improve gastrointestinal function. They also believed that preserving the pylorus would prevent exposure of the jejunum to excessive amounts of acid, thereby reducing the incidence of jejunal ulceration. Furthermore, they argued that resection of the distal stomach, pylorus, and proximal duodenum was unnecessary in the management of benign disease’. Indications and survival Traverso and Longmire’ initially performed P P P D in t w o patients. One had chronic pancreatitis with associated pseudocyst formation in the head of the pancreas while the 0007- 1323/90/090968- 07 zyxwv 0 1990 Butterworth-Heinemann Ltd zy z zyxwvutsrqponmlk Pylorous preserving pancreatoduodenectomy: P. A. Grace et al. Table 1 Indications Jor pyrfous preserving pancreatoduodenectomy zyxwvutsrqpo zy z zyxw Authors (reference) WatsonI6 Gall er a[.” Mosca et Flautner et a / . 2 3 Warshaw and TorchianaZ4 Pearlman et a[.25 Itani er a/.26 Crist et d Z 7 Eck and Murry” Kim et Braasch et a1.30-33 Williamson and Cooper34-36 Longmire et a/.5.3740 Hunt and McLean4’ McAfee et Sharp et Year No. of patients Chronic pancreatitis Malignancy Other 1944 1 0 1 1982 1984 18 30 18 0 0 0 0 37 0 1985 1986 1986 8 8 37 4 1 22 1985 I 1987 14 5 14 0 0 0 I 0 6 7 4 5 1987 2 1987 13 0 0 6 1988 126 41 4 82 1988 1988 1989 31 46 16 17 13 13 28 5 9 1989 1989 33 6 12 0 18 6 Total 394 163 (41) Values in parentheses are percentages second had a small localized carcinoma of the third part of the duodenum. Over 160 patients have since undergone PPPD Braasch and his for chronic pancreatitis (Table colleague^^^-^^ at the Lahey Clinic advocate the pylorus preserving operation in all resections of the right side of the pancreas for chronic pancreatitis except those patients with duodenal ulcer disease. However, the presence of pancreatitis throughout the gland requiring total pancreatectomy has also been combined with PPPD24.38.The advantages of pylorus preservation in these patients are that the patient can eat normal sized meals, and the absence of gastrectomy diminishes the risk of the dumping syndrome3. Doubts have been raised as to the advisability of performing the pylorus preserving operation for malignant disease in the periampullary area, as the field of resection may be reduced and survival thereby c ~ m p r o m i s e d ~Moossa4’ ~ * ~ ~ . has argued that the pylorus preservation procedure is not applicable to cancer of the head of the pancreas or distal common bile duct tumours since it may compromise the only chance of cure in these patients. Sharp and his colleagues43 have also reported tumour recurrence at the duodenal suture line. However, Newman and associates30 have observed that the other resection margins in pancreatoduodenectomy are much closer to the tumour than the duodenum and antrum. Performing frozen section of the duodenal resection margin should also minimize this problem. Therefore, leaving the antrum, pylorus, and first few centimetres of the duodenum is unlikely to compromise the field of resection. Moreover, several recent reports contain details of patients who underwent PPPD for a variety of periampullary malignant lesions including ductal pancreatic carcinoma with excellent long-term A total of 213 pancreatoduodenectomies have now been reported for malignant periampullary disease (Table I ) . Braasch and associate^^^^^^ have reported actuarial 5 year survival rates of 65 per cent for ampullary adenocarcinoma, 45 per cent for bile duct carcinoma, and 20 per cent for adenocarcinoma of the head of the pancreas following PPPD. These figures compare favourably with published survival rates following the standard Whipple resection: 37 per cent46 to 60 per cent47 for ampullary carcinoma; 17 per cent46 to 30 per cent3’ for carcinoma of the distal common bile duct; and 6 per ~ e n t to~ 18~per- centz7 ~ ~ for pancreatic adenocarcinoma. In a previous analysis from the University of California at Los Angeles, the present authors compared survival in a group Br. J. Surg., Vol. 77, No. 9, September 1990 2 213 (54) 0 3 3 1 0 zy 18 (5) of 26 patients with histologically proven adenocarcinoma of the head of the pancreas who were treated with either a PPPD ( n = 13) or a classical Whipple resection ( n = 13)39. The two groups of patients were comparable in terms of age, sex, preoperative laboratory data, and tumour staging. The actuarial survival rate was 25 per cent at 5 years for the pylorus preserving group compared with 12 per cent at 3 years for the classic Whipple group. Thus, patient survival appears not to be compromised following PPPD for adenocarcinoma of the head of the pancreas. We, therefore, agree with Braasch and co-authors who advocate PPPD in all patients with pancreatic and periampullary tumours except those cancers of the duodenum or bile duct that are in close proximity to the first centimetre of the d u o d e n ~ m ~ ~ . ~ ~ . Operative technique Adequate exposure is essential for assessment of resectability of the pancreas and subsequent pancreatoduodenectomy. The head of the pancreas is best approached through a right subcostal incision. This incision can easily be converted to a bilateral subcostal incision if a resection is to be undertaken. In a thin patient with a narrow costal angle, a midline upper abdominal incision also provides good exposure. In a patient with malignant disease, a full and comprehensive laparotomy is performed initially to detect metastases in the liver, lymph nodes, peritoneum, or omentum. A Kocher manoeuvre is performed and the head of the pancreas evaluated for size and location of the primary lesion. Invasion of contiguous structures such as the portal or superior mesenteric veins, the superior mesenteric artery, and the common hepatic artery can now also be assessed. If a preoperative histological diagnosis has not been made, biopsy of the primary lesion and any lymph nodes can be undertaken. Biopsy material can be obtained by open scalpel biopsy of any superficial lesion, transduodenal Tru-CutQneedle (Travenol Laboratories Inc., Deerfield, Illinois, USA) biopsy of the pancreatic mass or fine needle aspiration cytology of the lesion. The common duct may also be opened and curettage biopsy material obtained from the distal common bile duct. Controversy exists, however, as to whether histological proof of carcinoma of the head of the pancreas is necessary before embarking on pancreatoduodenectomy. Longmire and Traverse" have suggested that if the possibility 969 Pylorous preserving pancreatoduodenectomy: P. A. Grace et al. of a resection for benign disease has been discussed with the patient, it is acceptable to perform pancreatectomy without tissue diagnosis. Thus, a potentially curable occult lesion is not overlooked. Others have argued that if pancreatitis is severe enough to be confused with malignancy, pancreatectomy is probably indicated in any case4’. An assessment of resectability is now undertaken. The dissection is carried along the third part of the duodenum to expose the superior mesenteric vein at the point where it crosses the duodenum and passes under the neck of the pancreas. A critical determination of resectability is the ability to separate the portal and superior mesenteric veins from the adjacent pancreas. Major tributaries usually do not enter the portal vein on its ventral surface behind the pancreatic neck, and it should be possible to develop a plane between these two structures. If this manoeuvre can be achieved, resection is usually feasible. An intact neurovascular supply to the pylorus and the first part of the duodenum is essential for the success of the pylorus preserving operation. The right gastric artery is preserved, and the gastroduodenal artery is isolated and divided at the junction with the hepatic artery. The gastroepiploic vessels along the entire greater curvature of the stomach are carefully preserved as is the vagal innervation of the pylorus and antrum. The right gastroepiploic artery is divided at its origin from the anterior pancreaticoduodenal artery on the head of the pancreas (Figure I ) . The common hepatic duct should be divided just proximal to the cystic duct and a cholecystectomy performed. The duodenum is divided 2 cm distal to the pylorus. Unless a total pancreatectomy is being undertaken, the pancreas should be transected across the surgical neck. Transfixion sutures placed along the upper and lower margins of the pancreas on either side of the line of transection will reduce bleeding from the cut surface of the pancreas. Division of the small bowel distal to the ligament of Treitz will facilitate dissection of the uncinate process. Division of arterial and venous branches close to the jejunum at the ligament will help to avoid injury to the superior mesenteric artery. Bringing the divided jejunum to the patient’s right at this stage is also often a helpful manoeuvre. The uncinate process should now be carefully dissected from behind the superior mesenteric vein. Several small veins in this region must be individually ligated. The uncinate process may now be freed completely by dividing the remaining retroperitoneal tissues and branches to the superior mesenteric artery. zyxwvut Choledochojejunostomy zy zyxwvutsrqpo zyxwvuts Hepatic a r t e r y i ‘Duodenojejunostomy Paediatric feeding t u b e Figure 2 End-to-side duodenojejunostomy with the pancreaticojejunostomy and choledochoduodenostomy pluced proximally. A long paediatric feeding tube splints the pancreatojejunal anastomosis. (from Grace et zyxwv zyxwvuts zyxwvutsrqpon vesicle d i v i d e d Coeliac a r t e r LL ‘\ / Reconstructive methods Reconstruction of the alimentary tract is now undertaken. Two different methods of duodenojejunostomy have been described. Watson performed an end-to-end duodenojejunostomy in his operation 16, a technique also favoured by Flautner and colleague^^^, while Traverso and Longmire’ described an end-to-side anastomosis of the duodenal segment to a loop of jejunum. A variety of methods of managing the pancreatic stump have also been advocated. These include ligation of the pancreatic stumpL6,pancreatic duct occlu~ion’~, pancreaticojejunostomy’, and pancreaticoga~trostomy~~-~~. Pancreatic duct ligation is not to be recommended as it leads to a high incidence of fistula f ~ r m a t i o n ’ ~ . ’Pancreatic ~. duct occlusion also has the long-term problem of assuring pancreatic exocrine insufficiency and the need for enzyme replacement. To help neutralize gastric acid, the pancreaticojejunostomy and the choledochojejunostomy are placed proximal to the duodenojejunostomy. Placing the pancreaticojejunostomy proximal to the choledochojejunostomy also facilitates reoperation if resection is required for a leak at the pancreaticojejunostomy. A number of methods of pancreaticojejunostomy are available. If the pancreas is soft with a small pancreatic duct, as is often the case with malignant disease, the transected pancreatic stump can be invaginated into the proximal jejunum. This procedure is called the ‘dunking’ pancreaticojejunostomy and was originally described by Desjardins in 19079. However, if the duct is dilated and thickened from chronic pancreatitis, a mucosa-to-mucosa anastomosis between the pancreatic duct and the jejunum should be performed (Figure 2)54. Either an end-to-side or lateral pancreaticojejunostomy as described by Traverso and Longmire’ in their patient with pancreatitis can be undertaken. Stenting this anastomosis may reduce the incidence of postoperative pancreatic fistula38. Braasch and co-workers have indicated that in patients who d o not have chronic pancreatitis long-term patency of the pancreatic anastomosis may be better following the dunking procedure than after the end-to-side anastomosis3I . Recently, a number of authors have advocated pancreatogastrostomy as an ideal complement to PPPD for chronic p a n ~ r e a t i t i s ~ ~ ~ Flautner ”~”. and his colleagues23 argue that pancreatogastrostomy is technically easier than pancreatojejunostomy and that during the postoperative period the anastomosis may be easily examined by endoscopy or radiology. Furthermore, the alimentary tract is restored so that a blind intestinal segment does not remain. Recent evidence also suggests that pancreatogastrostomy may give better relief , zyxwvutsr zyxwvutsrqponml gastroepiploic aRight rtery divided A n t e r i o r a n d p o s t e r i o r pancreaticoduodenal a r t e r i e s Figure 1 Preseming the blood supply to the pylorus. The right gastric artery is preserved. The gustroduodenal artery is divided at its junction ivith the hepatic artery. The right gastroepiploic artery is dioided at its origin ,from the anterior pancreatoduodenal artery preserving the gastroepiploic vessels along the greater curvature of‘ the stomach. (from Grace et 970 Br. J. Surg.. Vol. 77, No. 9. September 1990 Pylorous preserving pancreatoduodenectorny: P. A. Grace e t al. of pain than pancreatojejunostomy in patients with chronic p a n ~ r e a t i t i s ~This ~ . procedure has the significant potential disadvantage, however, of leading to both a pancreatic and gastric fistula when an anastomotic leak occurs. Whenever possible, a small plastic catheter should be placed into the pancreatic duct to act as a splint for the pancreatic anastomosis. An end-to-side choledochojejunostomy is then created. Soft, closed suction drains are placed adjacent to the pancreatic and biliary anastomoses before closing the abdomen. z Hopkins also failed to observe any difference in the incidence of delayed gastric emptying between patients undergoing the pylorus preserving procedure or partial gastric resection (29 uersus 28 per cent)27.Similarly, McAfee et al.42from the Mayo Clinic found no differences between the PPPD and the classic Whipple operation with respect to mean days of nasogastric suction (7.6 uersus 7.9) or percentage of patients with early gastric stasis (28 uersus 33 per cent). Moreover, Hunt and McLean4’ have suggested that delayed gastric emptying in the early postoperative period is usually associated with another intra-abdominal complication. In the authors’ experience early Postoperative mortality delayed gastric emptying is most often associated with a leak In recent years, a substantial improvement in operative at the pancreatojejunal anastomosis or with postoperative mortality has been reported following pancreatoduodenectomy . pancreatitis. Twenty-two years ago, was able to report 41 Several investigators have studied gastric function in detail consecutive Whipple procedures without mortality. However, following pylorus preserving pancreatectomy in both humans his experience was unique as contemporary reports recorded and experimental animals. Traverso and Longmire3’, in a operative mortality rates of 20-40 per cent57-59. During the detailed study of gastric function using pyloric fluoroscopy and last decade, operative mortality rates decreased to 1&12 per gastric barium emptying, found normal gastric emptying in all cent in tertiary referral centre^^^,^^.^'. However, Herman in eight patients studied 2 to 12 months following pancreatic 197862, combining the results of 1005 Whipple procedures resection. Braasch and colleagues3 using isotopes to measure performed by 18 surgeons, reported an overall surgical gastric emptying of liquids and solids found no differences mortality rate of 20 per cent. Current reports suggest that between five control subjects and five patients following PPPD. pancreatoduodenectomy carries an operative mortality rate of Patti and c o - ~ o r k e r sin~ ~San Francisco measured gastric less than 5 per ~ e n t ~ ~ PPPD * ~ ~ is. ~associated ~ . ~ ~with . emptying and small bowel transit in ten patients following comparable operative mortality rates. Among 339 patients PPPD. They found that the rate of gastric emptying was normal in the recent literature, 13 perioperative deaths were in six patients, faster than normal in three, and abnormally reported for an operative mortality rate of 3.8 per cent slow in a single patient who was subsequently found to have (Table 2)21-28.g0-35*41-43. However, only two of the deaths diffuse carcinomatosis. Small bowel transit was found to be were procedure related. One patient died following dehiscence increased in seven patients, which Patti and his associates of the duodenojejunal anastomosis” while the second patient attributed to loss of exposure of the duodenum to food. died following a pancreatic fistula34.Furthermore, the reported Recent studies by Fink et al?’ by Hunt and McLean4’, and incidence of pancreatic f i s t ~ l a ~ ~biliary - ~ ~f .i ~ t~u l. a~~~’and .,~ ~ , by Cooper et a1.66 confirm that the majority of patients have enteric fistula” following PPPD is in the same range as after normal gastric emptying after PPPD. In comparison with the classic Whipple procedure. Moreover, these reports indicate patients undergoing the classic Whipple procedure, Cooper et that PPPD is a safe operation carrying with it a low operative a1.66found no significant differences. In the study by Fink and mortality. his ~ o l l e a g u e s gastric ~ ~ , emptying of liquids was normal after PPPD but delayed after the standard Whipple. Thus, of 58 patients without recurrent cancer who have had gastric Gastric and small bowel function emptying studies many months after surgery, only five One of the concerns expressed about the pylorus preserving (9 per cent) have had delayedemptying (Table 2)31*37.40*41.65.66. operation has been whether the stomach would empty Itani and colleagues67 studied the effects of duodenosatisfactorily following the procedure. In reviewing the jejunostomy on gastric emptying and on gastric motility in collected experience of 284 patients, transient delay in gastric dogs. These investigators found no significant difference in emptying was found to occur in 27.1 per cent of patients gastric emptying in either control or duodenojejunostomy animals during the 3-month period of observation. Similarly, (Table 2)22-24*27,30*31*34-38.39*41.42, Braasch and his colleagues3’ gastric myoelectric activity was unaltered by duodenonoted that 50 per cent of their patients required gastric suction for more than 7 days. Similarly, Warshaw and T ~ r c h i a n a ~ ~ jejunostomy. PPPD does not remove the natural duodenal pacemaker which is located approximately 0.5 cm distal to the observed a significant delay in achieving full diet in patients pylorus68. Altered small bowel motility has been demonstrated following the pylorus preserving procedure compared with the in patients following the classic pancreatodu~denectomy~~. classical Whipple procedure (16 days uersus 9.8 days). Mosca However, dumping and diarrhoea have not been long-term and his colleagues22 reported the resumption of normal problems following PPPD. gastrojejunal emptying around the tenth postoperative day, Despite normal gastric myoelectric activity, Itani and while 8 days was the longest period of nasogastric aspiration colleagues67 observed profound alterations in proximal small in Williamson and Cooper’s e ~ p e r i e n c e ~ ~ . bowel myoelectric activity following duodenojejunostomy in In a previous analysis, however, the present authors found dogs. Specifically, they observed a significant decrease in that postoperative nasogastric intubation was required for a pacesetter potential frequency in the bowel distal to the site of similar period in both pylorus preserving (6.7 days) and classical duodenal transection. The physiology of gastrointestinal Whipple (7.7 days) patientsg8. A recent study from Johns motility following PPPD remains to be completely elucidated. The role of hormones, gut peptides, and neurological factors Table 2 Postoperative mortality and morbidity rates in the regulation of gastrointestinal motility following pancreatoduodenectomy needs further study. The evidence to No. of date supports the view that in the majority of patients no patients Total significant long-term alterations in upper gastrointestinal Postoperative mortality rate 13 339 (3.8) motility occur following PPPD. Although delayed gastric emptying may be seen in a quarter of the patients in the early Delayed gastric emptying rate postoperative period, this problem is often related to a 77 284 (27.1) Postoperative pancreatic fistula or pancreatitis. Long-term 5 58 (9) A number of investigators have observed enterogastric reflux Marginal ulceration rate 12 331 (3.6) in some patients following PPPD. Braasch et aL3’ using 99mTc bound to iminodiacetic acid calculated an enterogastric reflux Values in parentheses are percentages zyxwvuts zyxwvu zyxwvutsrq zyxwvut ~~ ~~ ~ - zyxwvutsrqp zyxwvuts Br. J. Surg., Vol. 77, No. 9, September 1990 971 Pylorous preserving pancreatoduodenectomy: P. A. Grace et al. index which reflected the percentage of pancreatobiliary and jejunal contents that reflux into the stomach over a period following injection of the isotope. Although two of five patients showed evidence of reflux, they noted a wide divergence of reflux values following pyloric preservation. The other three patients studied had reflux curves similar to normal controls. Mosca and his colleagues2’ in an analysis of 14 patients post-pylorus preservation recorded biliogastric reflux in six patients. However, in only two patients was reflux marked and prolonged. Cooper et ~ 1 . ~ ‘have also documented that enterogastric reflux is more common after the classic Whipple procedure than after PPPD. B r a a ~ c hhas ~ ~even suggested that preservation of the pylorus by preventing alkaline reflux and antral stimulation may prevent peptic ulcer formation. However, the significance of reflux in patients following pylorus preservation is difficult to determine as it has been postulated that duodenogastric reflux is a physiological phenomenon in man”. Moreover, Fink and associates4’ found no mucosal lesions in their P P P D patients who underwent postoperative endoscopy. zyxwvut marginal ulceration demonstrated. Similarly, neither Flautner and his associates23, Perlman and his colleague^^^, nor Kim et a/.29reported marginal ulceration in a total of 60 patients. Crist and colleagues2’ reported a single case of upper gastrointestinal bleeding following PPPD among 14 patients (7 per cent). Warshaw and TorchianaZ4observed a single case ofjejunal ulceration in their series ofeight patients (12 per cent). This patient required antrectomy and vagotomy. McAfee et a/.4’ reported marginal ulceration in four of 31 patients (13 per cent). Hunt and McLean41 have reported the highest incidence of marginal ulceration with three of 16 patients (19 per cent) developing this problem. Thus, of 331 patients who have undergone PPPD, twelve (3.6 per cent) have had postoperative problems suggestive of marginal ulceration (Table 2)22-2 7 . 2 9 . 3 3 . 4 0 . The incidence of this problem has ranged from zero to 19 per cent which is in the same range as has been reported for the classic pancreatoduodenectomy which includes a vagotomy. Moreover, most patients with marginal ulceration after P P P D have been easily controlled with medical management. A number of authors have studied the effects of pancreatic resection on acid and gastrin production. Traverso and Longmire3’ studied basal and stimulated acid output in eight patients at a mean of six months following PPPD. They observed low-normal basal (0-4.1 mmol/h) and stimulated (0.1-224 mmol/h) acid production. Pearlman and his associates2’ also reported similar findings. Pearlman and associates reported a variable stimulated gastrin response after P P P D but persistent hypergastrinaemia was not seen28. Sudo and colleagues’’ analysed gastrin production before and after a standard Whipple resection. They found that both fasting and stimulated gastrin levels were reduced following pancreatoduodenectomy. Takuda and associate^'^ compared 20 patients with P P P D , 27 patients with the classic Whipple procedure, and eight healthy controls. They found that postprandial gastrin and secretin were normal after P P P D but significantly diminished after the classic Whipple. Similarly, Kim et a/.29found that serum gastrin and gastric acid were normal after their modification of the P P P D and decreased with the classic Whipple operation. Thus, both experimental and clinical evidence suggest that the PPPD is not an ulcerogenic procedure. zyxwvuts zyxwvutsrqpo zyxwvutsrqp Marginal ulceration Marginal ulceration has posed a considerable problem following classic pancreatoduodenectomy. Owens in 1948” reported three patients with peptic ulcer following Whipple operations and advised that the gastroenterostomy should be placed distal to the pancreatic and biliary anastomoses. Ten years later, Porter reported three marginal ulcers which occurred despite distal placement of the gastroenter~stomy’~. He advocated hemigastrectomy for prevention of the problem while Elliot73 argued that pancreatoduodenectomy inevitably led to gastric hyperacidity unless an antrectomy was performed. Grant and van Heerden” in an analysis of 297 patients who had undergone pancreatoduodenal resection at the Mayo Clinic observed anastomotic ulceration in 18 patients (6 per cent) of whom seven had undergone total pancreatectomy. These authors observed that total pancreatectomy was more ulcerogenic than a Whipple procedure and that a 50 per cent gastric resection was inadequate to prevent marginal ulceration. In an attempt to reduce the incidence of anastomotic ulceration, Scott and his colleague^'^ advocated truncal vagotomy with pancreatoduodenectomy. In an analysis of 41 pancreatoduodenectomies, they observed anastomotic ulceration in 36 per cent of patients without vagotomy compared with no ulceration in patients who had undergone vagotomy. However, Walsh and associate^'^ found that antrectomy alone was associated with a marginal ulcer rate of 12 per cent compared with 9 per cent when vagotomy was added. These findings have since been substantiated by Grace and associates38 and Crist and colleagues27 who found that the incidence of marginal ulceration was similar for antrectomy alone or antrectomy plus vagotomy. Watson’‘ and Traverso and Longmire’ believed that preserving the pylorus during pancreatic resection would decrease the postoperative incidence of jejunal ulceration. The clinical experience to date would support their hypothesis. Braasch and associates33 in their series of 126 patients reported six patients ( 5 per cent) in whom the diagnosis of marginal ulceration was suggested either radiologically o r by endoscopy. Four patients were managed successfully by medical means while two underwent operation. Grace et ul. observed no upper gastrointestinal bleeding in the immediate postoperative period in 39 patients following PPPD. but two patients ( 5 per cent) presented at 16 and 61 months after operation. One was found to have a marginal ulcer which required gastric resection while the other had non-specific upper gastrointestinal bleeding probably related to portal hypertension. In a later follow-up of this series by Fink et a/.40 two of 46 patients (5 per cent) had developed marginal ulceration. All 30 patients in the series by Mosca er a / . 2 2underwent upper gastrointestinal endoscopy. Five patients had moderate inflammation of the antral mucosa but in no patient was zyxwvutsr Postoperative nutrition One of the theoretical advantages of P P P D is improved nutrition. Most patients who undergo the classic Whipple procedure are unable to achieve their preoperative weight3’. However, Braasch and his colleague^^^^'^ have documented that patients undergoing P P P D regained a median of 101 per cent of preoperative weight and 95 per cent of preillness weight. Fink and his associates4’ also studied serum levels of vitamins A and B , 2 . iron, total iron-binding capacity, carotene. and folate in both P P P D and classic Whipple patients. N o differences were noted between the two groups, and vitamins A, B I Zand folate levels were normal in all patients. Of six patients undergoing PPPD, serum iron levels were low in two, total iron-binding capacity was increased in three, and serum carotene levels were low in five. However, similar abnormalities were observed in classic Whipple patients. Thus, current data suggest that weight and the majority of nutritional parameters are normal after PPPD. zyxwvut Conclusion P P P D represents one of the most significant recent advances in pancreatic surgery. The operation can be performed with a very low mortality and morbidity and is technically easier than the standard Whipple resection. Gastric emptying may be prolonged transiently in the immediate postoperative period. However, this complication is easily managed and is not a long-term problem. In addition, P P P D has a lower incidence of enterogastric reflux. dumping, and diarrhoea when compared with the classic Whipple procedure. Clinical and experimental zyxwvutsrqpo 972 Br. J. Surg., Vol. 77, No. 9, September 1990 zyx P y l o r o u s preserving p a n c r e a t o d u o d e n e c t o m y : P. A. G r a c e et al. evidence has substantiated the hypothesis that preservation of the pylorus would reduce the incidence of marginal ulceration following pancreatectomy. Moreover, patients undergoing PPPD are m o re likely t o regain their preoperative a n d preillness weight. Pancreatoduodenectomy with preservation of the pylorus may be the procedure of choice for chronic pancreatitis involving the head of the pancreas. Initial concerns a b o ut the use of this procedure in malignant disease have n o t been born out, a nd its use should be considered for curative or palliative resections of pancreatic and periampullary malignancies. Surgery is the only method of treatment which provides the chance of cure in patients with such pancreatic an d periampullary neoplasms, and may also provide the best palliation in terms of survival and quality of life. Thus, we believe that PPPD has a significant role to play in the management of patients with pancreatic a n d periampullary neoplasms. 24. 25. 26. 27. zyxwvutsrqpo zyxwvutsrq zyxwvutsr zyxwvutsrqpo zyxwvutsrqpo zyxw zyxwvu 28. 29. 30. 31. References I. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. of the pylorus in the treatment ofchronic pancreatitis. Am JSur,q 1985; 150: 608-1 1. Warshaw AL, Torchiana DL. Delayed gastric emptying after pylorus preserving pancreaticoduodenectomy. Surg Gynecol Ohstet 1985; 160: 1 4 . Pearlman NW, Stiegmann GV, Ahnen DJ, Schultz AL, Fink LM. Acid and gastrin levels following pyloric-preserving pancreaticoduodenectomy. Arch Surg 1986; 121: 6 6 1 4 . Itani KMF, Coleman RE, Akwari OE, Meyers WC. Pylorus preserving pancreatoduodenectomy. Ann Sury 1986;204: 655-64. Crist DW, Sitzmann JV, Cameron JL. Improved hospital morbidity, mortality and survival after the Whipple procedure. Ann Surg 1987; 206: 358-65. Eck G, Murry JW. Pyloric preservation with pancreaticoduodenectomy. South Med J 1987; 80: 1048-50. Kim H-C, Suzuki T, Kajiwara T, Miyashita T, Imamura M , Tobe T. Exocrine and endocrine stomach after gastrobulbar preserving pancreatoduodenectomy. Ann Surg 1987;206: 71 7-27. Newman KD, Braasch JW, Rossi RL, O’Campo-Gonzales S. Pyloric and gastric preservation with pancreatoduodenectomy. A m J Surg 1983; 145: 152-6. Braasch JW, Rossi RL, Watkins E. Deziel DJ, Winter PF. Pyloric and gastric preserving pancreatic resection. Ann Surg 1986; 204: 41 1-8. Braasch JW. Pancreaticoduodenal resection. Curr Prohl Surg 1988; 25: 323-63. Martin FM, Rossi RL, Braasch JW. Pylorus-preserving pancreatoduodenectomy. Results with 126 consecutive patients. Neth JSurg 1988; World Congress on HPB Surgery (Suppl): 194. Cooper MJ, Williamson RCN. Conservative pancreatectomy. Br J Surg 1985; 72: 801-3. Williamson RCN, Cooper MJ. Resection in chronic pancreatitis. Br J Surg 1987; 74: 807-12. Williamson RCN, Cooper MJ. Resection of the pancreatic head. Nerh JSurg 1988:World Congress ofHPB Surgery (Suppl):26. Traverso LW, Longmire WP jun. Preservation of the pylorus in pancreaticoduodenectomy. A follow up evaluation. Ann Surg 1980; 192: 306-10. Grace PA, Pitt HA, Tompkins RK, DenBesten L, Longmire WP jun. Decreased morbidity and mortality after pancreatoduodenectomy. Am J Surg 1986; 151: 141-9. Grace PA, Pitt HA, Longmire WP. Pancreatoduodenectomy with pylorus preservation for adenocarcinoma of the head of the pancreas. Br J Surg 1986; 73: 647-50. Fink AS, Desouza LR, Mayer EA, Hawkins R, Longmire WP. Long-term evaluation of pylorus preservation during pancreatoduodenectomy. World J Surg 1988; 12: 663-70. Hunt DR, McLean R. Pylorus-preserving pancreatectomy: functional results. Br J Surg 1989; 76: 173-6. McAfee MK, van Heerden JA, Adson MA. Is proximal pancreatoduodenectomy with pyloric preservation superior to total pancreatectomy? Surgery 1989; 105: 347-51. Sharp KW, Ross CB, Halter SA et a/. Pancreatoduodenectomy with pyloric preservation for carcinoma of the pancreas: a cautionary note. Surgery 1989; 105: 645-53. Hayes DH, Bolton JS, Willis GW. Bowen JC. Carcinoma of the ampulla of Vater. Ann Surg 1987; 206: 572-7. Moossa AR. Pancreatic cancer. Approach to diagnosis, selection for surgery and choice of operation. Cancer 1982; 50: 2689-98. Tarazi RY, Hermann RE, Vogt DP et a/. Results of surgical treatment of periampullary tumours: 35 year experience. Surgery 1986; 100: 716-22. Neoptolemos JP, Talbot IC, Carr-Locke DL e t a / .Treatment and outcome in 52 consecutive cases of ampullary carcinoma. Br J Surg 1987; 74: 957-61. Connolly MM, Dawson PJ, Michelassi F, Moosa AR, Lowenstein F. Survival in 1001 patients with carcinoma of the pancreas. Ann Surg 1987; 206: 366-73. Howard JM, Jordan GL. Cancer of the pancreas. Curr Prohl Cancer 1977; 2: 5-52. Flautner L, Tihanyi T, Bock G, Mako E, Szecseny A. A new approach - pancreatoduodenectomy with preservation of the pylorus in the surgical treatment of diseaes of the pancreas head. Orv Heti/ 1982; 123: 1181-3. lcard P, Dubois F. Pancreaticogastrostomy following pancreatoduodenectomy. Ann Surg 1988; 207: 253-6. Aston SJ. Longmire WP. Pancreaticoduodenal resection: 20 years experience. Arch Surg 1973; 106: 813-7. Papachristou DN, D’Agostino H, Fortner JG. Ligation of the pancreatic duct in pancreatectomy. Br J Surg 1980; 67: 2-2. Silverberg E. American Cancer Society. Cancer Srati.stic.7 1986; 36: 9-16. Gyr K, Heitz PU, Beglinger C. Pancreatitis. In: Kloppel G, Heitz PU, eds. Panereuric Parliology. Edinburgh: Churchill Livingstone, 1948: 4-72. Moossa AR. Surgical treatment of chronic pancreatitis: an overview. Br J Surg 1987; 74: 661-7. Fortner JG. Regional pancreatectomy for cancer of the pancreas, ampulla and other related sites. Ann Surg 1984; 199: 418-25. Traverso LW, Longmire WP jun. Preservation of the pylorus in pancreaticoduodenectomy. Surg Gynecol Ohsrer 1978; 146: 959-62. Lambert MA, Lincham IP. Russell RCG. Duodenum preserving total pancreatectomy for end stage chronic pancreatitis. Br J Sury 1987; 74: 35-9. Knox RA, Kingston RD. Carcinoma of the ampulla of Vater. Br J Surg 1986; 73: 72-3. Halsted WS. Contributions to the surgery of the bile passages. especially of the common bile duct. Boston Med Surg J 1899; 141: 645-54. Desjardins A. Technique de la pancreatectomie. Rev Cliir 1907; 35: 945-73. Sauve L. Des pancreatectomies et specialement de la pancreatectomie cephalique. Rer Chir 1908; 37: 335-85. Kaunsch W. Das carcinoma der pappilla duodeni und seine radikale entfernung. Beitrage Z Klin Chir 1912; 78: 439-86. Tenani 0. Contribution to the surgery of the papilla of Vater. Policlinico 1922; 29: 291-300. Whipple AO, Parsons WB, Mullens CR. Treatment ofcarcinoma of the ampulla of Vater. Ann Sury 1935; 102: 763-79. Trimble IR. Parsons JW, Sherman CP. A one stage operation for the cure of carcinoma of the ampulla of Vater and the head of the pancreas. Surg Gynecol Ohster 1941; 73: 71 1-22. Whipple AO. The rationale for cancer of the pancreas and ampullary region. Ann Surg 1941; 114: 612-15. Watson K. Carcinoma of the ampulla of Vater. Successful radical resection. Br J Surg 1944: 31: 368-73. Warren KW, Veidenheimer MC, Pratt HS. Pancreaticoduodenectomy for periampullary cancer. Sury Clin North .4m 1967; 47: 63945. Grant CS, van Heerden JA. Anastomotic ulceration following subtotal and total pancreatectomy. Anri Surg 1979; 190: 1-5. Scott JW, Dean RH, Parker T. Avant G. The role of vagotomy in pancreaticoduodenectomy. Ann Surg 1980: 191 : 688-96. Longmire WP jun., Traverso WL. The Whipple procedure and other standard operative approaches to pancreatic cancer. Cancer 1981;47: 1706-11. Gall FP. Gebhardt C, Zirngibl H. Chronic pancreatitis: results in I 16 consecutive, partial duodenopancreatectomies combined with pancreatic duct occlusion. Hepa/ogastroentero/og~~ 1982; 29: 115-9. Mosca F, Guilianotti PC. Arganini M, Brandi LS. Selli M. Pancreaticoduodenectomy with pylorus preservation. I t J Surg Sci 1984; 14: 3 13-20, Flautner L, Tihanyi T, Szecseny A. Pancreatogastrostomy: an ideal complement to pancreatic head resection with preservation Br. J. Surg., Vol. 77, No. 9, S e p t e m b e r 1990 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 973 zyxwvu zyxwvut zyxwvutsrqp zyxw Pylorous preserving pancreatoduodenectomy: P. A. Grace et al. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 974 Cattell RB. A technique for pancreatoduodenal resection. Surg Clin North A m 1948; 28: 761-75. Pain JA, Knight MJ. Pancreaticogastrostomy: the preferred operation for pain relief in chronic pancreatitis. Br J Surg 1988; 75: 2 2 c 2 . Howard JM. Pancreaticoduodenectomy:41 consecutive Whipple resections without an operative mortality. Ann Surg 1968; 168: 62940. Morris PJ, Nardi GL. Pancreaticoduodenal cancer. Arch Surg 1966; 92: 83+7. Lansing PB, Blalock JB, Oschner JC. Pancreaticoduodenectomy: retrospective review 1949-1969. A m J Surg 1972; 38: 79-86. Gilsdorf RB, Spanos P. Factors influencing morbidity and mortality in pancreaticoduodenectomy. Ann Surg 1973; 177: 332-7. Warren KW, Choe DS, Plaza J, Relihan M. Results of radical resection for periampullary cancer. Ann Surg 1975; 181: 5 3 4 4 0 . Forrest J, Longmire WP jun. Carcinoma of the pancreas and periampullary region. A study of 279 patients. Ann Surg 1979; 189: 129-38. Hermann RE. Panel: cancer of the pancreas. A m J Surg 1978; 135: 185-91. Trede M. The surgical treatment of pancreatic carcinoma. Surgery 1985; 97: 28-35. Jones BA, Langer B, Taylor BR, Girotti M. Periampullary tumours: which ones should be resected? A m J Surg 1985; 149: 4652. Patti MG, Pellegrini CA, Way LW. Gastric emptying and small bowel transit of solid food after pylorus preserving pancreaticoduodenectomy. Arch Surg 1987; 122: 528-32. Cooper MJ, Rhys-Davies E, Williams RCN. Gastric function following conventional and pylorus preserving pancreato- 67. 68. 69. 70. 71. 72. 73. 74. 75. duodenectomy. N e f h J Surg 1988; World Congress of HPB Surgery (Suppl): 242. Itani KMF, Coleman E, Akwari OE. Pylorus preserving pancreatoduodenectomy : does duodenojejunostomy alter motility and gastric emptying? Surg Forum 1986; 37: 195-7. Hermon-Taylor JH, Code CF. Localization of the duodenal pacemaker and its role in the organization of duodenal myoelectric activity. Cur 1971; 12: 4 c 7 . Dahlgren S, Selking 0. Postoperaive motility of the small intestine: a study with endoradiosondes. Ups J Med Sci 1972; 77: 2 0 2 4 . Fisher RS, Cohen S. Pyloric sphincter dysfunction in patients with gastric ulcer. N Engl J Med 1973; 288: 2 7 3 4 . Owens FM sen. Problem of peptic ulcer following pancreatectomy. Ann Surg 1948; 128: 15-20. Porter MR. Carcinoma of the pancreaticoduodenal area: operability and choice ofprocedure. Ann Surg 1958; 148: 71 1-24. Elliott DW. Emect of lost pancreatic juice on gastric and peptic ulcer. A m J Surg 1974; 127: 6 5 8 4 2 . Walsh DB, Eckhauser FE, Cronenwett J, Turcotte JG, Lindenauer SM. Adenocarcinoma of the ampulla of Vater: diagnosis and treatment. Ann Surg 1982; 195: 152-7. Sudo T, Ishiyama K, Kawamura M ef al. Changes in plasma gastrin and secretin levels after pancreaticoduodenectomy. Surg Gynecol Obstet 1984; 158: 1 3 3 4 . Takada T, Yasuda H, Shikata J-I, Wdtanabe S-I, Shiratori K, Takeachi T. Postprandial plasma gastrin and secretin concentrations after a pancreatoduodenectomy. A comparison between a pylorus-preserving pancreatoduodenectomy and the Whipple procedure. Ann Surg 1989; 210: 47-51. zyxwvutsrqponm 76. Paper accepted 9 April 1990 Br. J. Surg., Vol. 77, No. 9. September 1990