Search PubMed⌕ Search

Biomedical subjects

M G Sarr

Publications and source records attributed to M G Sarr.

At least 127 records · Page 7Linked to original sources
PubMed1995

Incidence and management of pancreatic and enteric fistulas after surgical management of severe necrotizing pancreatitis.

OBJECTIVE: To determine the incidence, type, and outcome of complications of necrotizing pancreatitis. SETTING: Major tertiary referral center (Mayo Clinic, Rochester, Minn). PATIENTS: Sixty-one patients seen from 1985 to 1994 who underwent surgical management of severe necrotizing pancreatitis and who developed pancreatic or gastrointestinal fistulas. MAIN OUTCOME MEASURES: Incidence, management, and outcome of pancreatic and gastrointestinal fistulas. RESULTS: Twenty-five patients (41%) developed pancreatic (14 patients) and/or gastrointestinal tract cutaneous (19 patients) fistulas. While three duodenal fistulas and one colonic fistula were recognized at the initial operation for pancreatic necrosectomy, the remainder developed 4 to 60 days after the initial operation. Spontaneous closure occurred in nine of 14 pancreatic, two of two gastric, two of four enteric, two of eight colonic, and four of five duodenal fistulas. Mortality of the group with fistulas was 24% (6/25) and was not different from the mortality of the patients with necrotizing pancreatitis without fistulas (28% [10/36]). CONCLUSIONS: Pancreatic and gastrointestinal tract fistulas are common complications of surgical treatment of severe necrotizing pancreatitis. Well-controlled gastric, pancreatic, and enteric fistulas have the greatest likelihood of spontaneous closure. Duodenal and colonic fistulas may need surgical intervention for control or repair. Mortality in these patients parallels the mortality for severe necrotizing pancreatitis.

Colonic Diseases↗
PubMed1995

Technical and practical considerations involved in operations on patients weighing more than 270 kg.

Care of the patient with superobesity requires special precautions and appropriate equipment. Recently, we performed bariatric procedures (modified very-long-limb Rouxen-Y gastric bypass) on patients weighing 355 kg and 377 kg. These procedures required preoperative preparation concerning safe means of transport of the patient, techniques of anesthesia and intraoperative exposure, provisions for postoperative recovery, and measures to assure patient comfort and hygiene postoperatively. In addition to specially designed bariatric procedures for the superobese, specialized equipment is needed to protect the health of the patient and the staff. All health care providers and especially acute care centers must have preconceived protocols to treat the superobese patient. In addition, specialized equipment is necessary to allow safe transport and support of these patients.

Adult↗
PubMed1995

Selective operative cholangiography. Appropriate management for laparoscopic cholecystectomy.

OBJECTIVE: To evaluate the results of selective intraoperative cholangiography (IOC) in patients undergoing laparoscopic cholecystectomy. DESIGN: Retrospective study. SETTING: Mayo Clinic, Rochester, Minn, from 1990 to 1991. PATIENTS: Five hundred forty-two patients underwent attempted laparoscopic cholecystectomy. Excluding 28 (5.2%) who underwent conversion to laparotomy and 19 (3.5%) who did not respond to a follow-up questionnaire, there were 495 respondents (mean follow-up, 25 months). MAIN OUTCOME MEASURE: Incidence and management of choledocholithiasis, extrahepatic bile duct injuries, and other findings potentially affected by IOC. RESULTS: Twenty patients underwent preoperative endoscopic retrograde cholangiopancreatography for suspected common bile duct abnormalities, and 10 had common bile duct stones removed. Nearly a third (n = 161 [32.5%]) of the patients underwent IOC for laboratory, historical, or operative findings or for training purposes. Common bile duct stones were discovered on IOC in five patients (3.1%), three of whom were treated successfully with postoperative endoscopic therapy; the two others had normal findings on endoscopic retrograde cholangiopancreatography (false-positive results of IOC). In three other patients in whom IOC was unsuccessful or incomplete, symptomatic common bile duct stones developed. Two patients were treated with endoscopic techniques, and one required open common bile duct exploration. Among the 334 patients who did not undergo IOC, symptoms suggestive of retained stones developed in eight (2.4%) (all within 2 months of surgery; mean, 18 days), but stones were found at endoscopy retrograde cholangiopancreatography in only four patients. Two had preoperative criteria for performing IOC. In only three patients (0.6%) from the study population would symptomatic retained common bile duct stones have developed with selective IOC and routinely successful IOC. No common bile duct injuries occurred. CONCLUSIONS: Selective IOC during laparoscopic cholecystectomy is a safe practice when the ductal anatomy is clearly defined and there is no laboratory or clinical evidence of common bile duct abnormalities. Symptomatic retained common bile duct stones will be infrequent, and bile duct injuries will be rare when IOC is performed for the appropriate indications. These data do not support the need for routine IOC, although this procedure is an essential tool for the laparoscopic surgeon.

Adult↗
PubMed1995

Percutaneous cholecystostomy: a valuable technique in high-risk patients with presumed acute cholecystitis.

Percutaneous cholecystostomy offers a potentially important therapeutic modality for critically ill patients with acute cholecystitis who represent a high risk for general anaesthesia. The aim of the study was to assess experience with percutaneous cholecystostomy in resolving the acute episode of cholecystitis without operative intervention. Twenty-two consecutive patients with a clinical diagnosis of acute cholecystitis underwent the procedure. All were at high risk for general anaesthesia, and all but one developed cholecystitis while hospitalized for another co-morbid condition; 14 were in an intensive care unit. Twenty-one of the 22 patients proved to have acute cholecystitis (11 acalculous, ten cholelithiasis). There were no acute technical complications. Toxaemia resolved in 17 of the 21 patients with acute cholecystitis. Acute cholecystitis failed to resolve in three patients; all died within 48 h from overwhelming generalized sepsis. One patient required emergency cholecystectomy for bile peritonitis when the cholecystostomy catheter became dislodged 24 h after placement. The 60-day mortality rate for the acalculous and calculous patient groups was 55 and 20 per cent, respectively. Only three interval cholecystectomies have been performed at a mean follow-up of 19 months. In conclusion, percutaneous cholecystostomy may be the procedure of choice for the management of acute cholecystitis in the very high-risk critically ill patient. If symptoms fail to resolve quickly, ongoing sepsis, cholangitis or gallbladder necrosis should be suspected.

Acute Disease↗
PubMed1995

Role of the vagal branches to the proximal stomach in mediating gastric distention-induced disruption of canine interdigestive upper gut motility.

Previous experiments from our laboratory have shown that the vagus nerves mediate proximal gastric distention-induced disruption of interdigestive motor patterns in the upper gut of dogs. Our aim was to determine the role of vagal innervation of the proximal stomach in mediating the response to nonnutrient proximal gastric distention. Five dogs underwent proximal gastric vagotomy (PGV) and placement of electrodes and manometry catheters on the antrum and the upper small intestine. Proximal gastric distention for 5 hr was achieved by inflating a thin, compliant bag in the proximal stomach. Four volumes of distention stimulus (0, 1.5, 12.5, and 25 ml/kg) were tested. As with total abdominal vagotomy, intragastric stimulus volumes of 12.5 and 25 ml/kg after PGV no longer inhibited cycling of the migrating motor complex in the stomach, duodenum, proximal jejunum, and distal jejunum. Volumes of 12.5 and 25 ml/kg did, however, on occasion, lead to the absence of phase III activity in the stomach or the duodenum when it would have been expected to precede phase III activity in the jejunum; this effect did not occur in the jejunum. These findings with a nonnutrient stimulus suggest that vagal branches to the proximal stomach might mediate, in part, the postprandial changes in upper gut motility in response to gastric distention by ingestion of a meal.

Animals↗
PubMed1995

Effect of enteric nonnutrient infusions on motor patterns in neurally intact and neurally isolated canine jejunum.

Previous work in our laboratory has shown that nonnutrient mechanical factors initiate changes in motility patterns in local and remote regions of the small intestine. Our aims were to determine how isolated duodenal and jejunoileal nonnutrient infusions alter interdigestive motor patterns locally and distantly and whether these effects are neurally mediated. Ten dogs were prepared with duodenal and proximal jejunal infusion and manometry catheters and a proximal jejunal diverting cannula. Five of these dogs served as neurally intact controls (Group 1) and five also underwent in situ neural isolation of the entire jejunoileum (Group 2: extrinsic denervation; disruption of enteric myoneural continuity with duodenum). After recovery, nonnutrient infusions at 0-15 ml/min for 5 hr into proximal duodenum or jejunum did not consistently affect cycling of the migrating motor complex (MMC). The period and duration of individual phases of the MMC and time to first phase III after the start of infusion were similar in both groups. In Group 2, duodenal characteristics (period and duration of phase II, time to first phase III) increased slightly with increasing rates of jejunal but not duodenal infusion. Motility indices, although greater in Group 2, were not altered by enteric infusions. Differing rates of nonnutrient enteric flow limited to duodenum or jejunoileum did not affect markedly local or distant motor patterns. Alterations in interdigestive motility patterns by postprandial nonnutrient intraluminal content are not mediated directly by intraluminal flow.

Animals↗
PubMed1995

Calcium currents in human and canine jejunal circular smooth muscle cells.

BACKGROUND & AIMS: Although calcium plays an essential role in intestinal smooth muscle contractile activity, calcium entry pathways in canine and human small intestine are largely unknown. The goal of this study was to characterize calcium channels, a potential entry pathway for calcium, in isolated circular smooth muscle cells of canine and human jejunum. METHODS: Single freshly dissociated human and canine jejunal circular smooth muscle cells were studied using single-channel and perforated whole-cell patch clamp recordings as well as fluorescence dual wavelength ratio imaging. RESULTS: An inward whole-cell current was identified that was carried by a 17 pS (80 mmol/L Ba2+) dihydropyridine-sensitive, barium-permeable channel. The current was potentiated by BayK 8644 (1 mumol/L; n = 3; 82% +/- 34%), acetylcholine (1 mumol/L; n = 8; 42% +/- 5%), and erythromycin (1 mumol/L; n = 9; 70% +/- 11%) and was completely blocked by nifedipine (1 mumol/L; n = 6) or diltiazem (200 mumol/L; n = 4). Application of BayK 8644 (1 mumol/L), acetylcholine (1 mumol/L), or erythromycin (1 mumol/L) to Fura-2-loaded smooth muscle cells bathed in Krebs' solution containing 2.54 mmol/L calcium increased intracellular calcium levels. CONCLUSIONS: A calcium entry pathway was identified in canine and human jejunal circular smooth muscle cells. The pathway was mediated by a dihydropyridine-sensitive calcium channel. The channel allowed the entry of significant amounts of calcium at physiological extracellular calcium concentration.

3-Pyridinecarboxylic acid, 1,4-dihydro-2,6-dimethy↗
PubMed1995

Intraileal carbohydrate regulates canine postprandial pancreaticobiliary secretion and upper gut motility.

BACKGROUND & AIMS: The effect of nutrients in the distal small intestine or colon on postprandial upper gut function is incompletely understood. The aim of this study was to determine if carbohydrate in the ileum or proximal colon of dogs affects postprandial pancreaticobiliary secretion, gastrointestinal transit, and circulating concentrations of certain gastrointestinal regulatory peptides. METHODS: Seven dogs were prepared with permanent infusion and aspiration catheters in the duodenum and ileum and an infusion catheter in the cecum. Coincident with eating a meal containing liquid and solid markers, ileal or colonic (n = 5 dogs for each) infusion were begun of isosmolar 0.9% NaCl or carbohydrate in a 3:1 ratio of starch to glucose. Pancreatic enzyme output, bile acid delivery, gastrointestinal polypeptide, and plasma concentrations of pancreatic polypeptide, neurotensin, and peptide YY were measured for 6 hours postprandially. RESULTS: Carbohydrate infusion in the ileum, but not in the proximal colon, increased amylase secretion and plasma peptide YY, slowed gastric emptying of liquids and solids, slowed small intestinal transit, and decreased bile acid delivery into the duodenum (P < 0.05 in each). CONCLUSIONS: Carbohydrate in the ileum regulates postprandial exocrine pancreatic enzyme secretion and other postprandial upper gut functions. Peptide YY may play a role in this regulation.

Amylases↗
PubMed1995

Early clinical results with the uncut Roux reconstruction after gastrectomy: limitations of the stapling technique.

BACKGROUND: The aim of this study was to review our early clinical experience with the uncut Roux gastrectomy, a modified Billroth II gastrojejunostomy in which four rows of staples occlude the afferent jejunal lumen, while biliary and pancreatic secretions are diverted distally through a jejunojejunostomy. METHODS: Between September 1991 and April 1993, 9 women and 5 men underwent uncut Roux gastrectomy for gastric adenocarcinoma (5), postvagotomy or idiopathic gastroparesis (7), Roux stasis syndrome (1) or anastomotic ulceration with gastric outlet obstruction (1). Subsequently, patients were seen or contacted by phone within the 6 months prior to March 1994. RESULTS: Eight of the 14 patients (57%) had excellent results with stable weight and no nausea, vomiting, heartburn, abdominal pain, or postprandial symptoms. One patient continued to have early satiety and vomiting but maintained stable weight for 19 months postoperatively. Five patients (36%) had poor results with alkaline reflux gastritis or esophagitis. All 5 had documented staple line dehiscence. Four of them were reoperated on and converted to a standard Roux operation with relief of symptoms. CONCLUSIONS: The uncut Roux operation prevents the Roux stasis syndrome, but the current technique has an unacceptably high incidence of dehiscence of the staple lines with subsequent reflux gastritis or esophagitis.

Adult↗
PubMed1995

Intestinal transplantation: effects on ileal enteric absorptive physiology.

BACKGROUND: The effects of small intestine transplantation on enteric physiology are poorly understood. After orthotopic jejunoileal autotransplantation, dogs develop a severe watery diarrhea and lose up to 15% of their body weight. The cause of these changes has not been explained. Our aim was to determine the influence of jejunoileal autotransplantation on ileal absorption of water, electrolytes, and bile salts and the effects of proabsorptive and prosecretory agents on ileal transport. METHODS: Seven dogs were studied before and at 2 and 8 weeks after in situ jejunoileal neural and lymphatic isolation (a model of small intestine autotransplantation). With a triple-lumen perfusion technique, net ileal fluxes of water, electrolytes, and bile salts were measured before and at 2 and 8 weeks after this model of jejunoileal autotransplantation. In addition, the effects of an intravenous infusion of vasoactive intestinal polypeptide (a prosecretory agent) and norepinephrine (a proabsorptive agent) on net transport were evaluated. RESULTS: Dogs developed a profuse diarrhea after this model of autotransplantation. Ileal absorption of water and electrolytes decreased immediately (measured during operation), remained decreased for 2 weeks, and returned toward baseline by 8 weeks. A similar decrease in net flux of bile salts was shown at 2 weeks after transplantation and returned toward baseline by 8 weeks. The prosecretory response of vasoactive intestinal polypeptide on ileal fluxes of water and electrolytes was unchanged, whereas the proabsorptive response to norepinephrine increased after this model of autotransplantation. CONCLUSIONS: Jejunoileal autotransplantation decreases ileal absorption of water, electrolytes, and bile salts. The profuse watery diarrhea observed in dogs after small intestine autotransplantation may be a secretory and/or a bile salt-induced diarrhea related to the effects of jejunoileal denervation.

Animals↗
PubMed1995

Long-term survival after resection for ductal adenocarcinoma of the pancreas. Is it really improving?

OBJECTIVE: The authors review their recent experience with resected pancreatic ductal adenocarcinoma. SUMMARY BACKGROUND DATA: Ductal adenocarcinoma of the pancreas has traditionally had a 5-year survival rate less than 10% after curative resection. Recently, several groups have reported markedly improved 5-year survival rates (approaching 25%) for patients undergoing curative resection. METHODS: Institutional experience with 186 consecutive patients (1981-1991) with pathologic diagnoses of ductal adenocarcinoma undergoing pancreatic resection was reviewed. Histologic specimens of all 3-year survivors (n = 31) were re-reviewed by two pathologists, one internal and one external; nonductal pancreatic cancers then were excluded. RESULTS: After histologic re-review, 12 patients did not have ductal adenocarcinoma, leaving a total of 174 patients for analysis (102 men, 72 women; mean age 63 years, range 34-82 years). Mean follow-up was 22 months (range 4-109). Classical pancreaticoduodenectomy was performed in 71%, pylorus-preserving resection in 9%, and total pancreatectomy in 20%. Hospital mortality was 3%. Twenty-eight patients (16%) had macroscopically incomplete resections; 98 (56%) had lymph node metastases within the resected specimens, and 21 patients (12%) had extensive perineural invasion. Overall actuarial 5-year survival was 6.8%. Five-year survival was greater for node-negative versus node-positive patients (14% vs. 1%, p < 0.001), and for smaller (< 2 cm) versus larger tumors (20% vs. 1%, p < 0.001). The 5-year survival for the subset of patients with negative nodes and no perineural or duodenal invasion (69 patients) was 23% (p < 0.001). Mean survival of the 12 excluded patients was 53 +/- 7 months compared with 17.5 +/- 1 months in the 174 patients with ductal pancreatic cancer. CONCLUSIONS: Five-year survival for patients undergoing pancreatic resection for lesions deemed to be clinically "curable" intraoperatively and histologically reviewed/confirmed to be ductal adenocarcinoma of the pancreas is approximately 7%. Survival is greater (23%) in the subset of patients with negative nodes and no duodenal or perineural invasions. Pathologic review of all patients with pancreatic ductal cancer adenocarcinoma is mandatory if survival data are to be meaningful.

Adult↗
PubMed1995

An argument against routine percutaneous biopsy, ERCP, or biliary stent placement in patients with clinically resectable periampullary masses: a surgical perspective.

Improve resolution of computed tomography (CT) and ultrasonography allows us to visualize the proximal extent of biliary obstruction and the presence of a periampullary mass in most patients with malignant extrahepatic biliary obstruction. Our purpose in this report is to challenge the need for preoperative percutaneous biopsy, endoscopic retrograde cholangiopancreatography, or preoperative placement of a biliary endoprosthesis in the good-risk patient in whom the imaging procedure clearly defines a periampullary mass and the proximal extent (hepatic extent) of biliary obstruction. We recently managed three patients in whom one of these invasive procedures led to a complication that delayed, prevented, or complicated appropriate operative resection of a pancreatic neoplasm. Because a negative percutaneous biopsy, cholangiographic imaging of a dilated bile/pancreatic duct clearly seen on CT or ultrasonography, or short-term preoperative biliary decompression does not alter the decision for operative exploration and may cause complications, we argue against their use in the good-risk patient with both extrahepatic biliary obstruction and a periampullary pancreatic mass well delineated on noninvasive imaging.

Aged↗
PubMed1995

PYY and GLP-1 contribute to feedback inhibition from the canine ileum and colon.

To explore mechanisms whereby unabsorbed nutrients in the ileum inhibit the upper gut ("ileal brake"), we perfused the canine ileum or colon and monitored phase 3 in the duodenum. Fasting motility was recorded when the ileum or colon was perfused with 154 mM NaCl, a mixed isotonic nutrient solution (Ensure), or individual nutrients (maltose, casein hydrolysates, or sodium oleate). Blood samples were collected before and during the perfusions. The ileum was also perfused with 154 mM NaCl while peptide YY (PYY) was infused by vein. In both sets of experiments, plasma levels of PYY, neurotensin, and glucagon-like peptide-1 (GLP-1) were measured. Ileal or colonic perfusion of Ensure delayed phase 3 [migrating motor complexes (MMC)] in the duodenum, inhibited ileal motility, and increased plasma levels of PYY and GLP-1. Ileal casein and oleate and colonic casein also delayed the duodenal MMC. The MMC cycle length and plasma levels of PYY were closely correlated. Intravenous PYY prolonged the MMC cycle; an intravenous dose of 100 pmol.kg-1.h-1 of PYY mimicked the effects of ileal Ensure. These results support the hypothesis that PYY, and possibly GLP-1, participate in the ileal brake. This negative feedback loop also affects the distal small bowel. The proximal colon also triggers the feedback inhibition of gut motility (colonic brake).

Animal Nutritional Physiological Phenomena↗
PubMed1995

General surgical laparoscopic procedures for the "nonlaparologist".

OBJECTIVE: To assess our initial experience with a variety of laparoscopic procedures that can be performed by general surgeons with the basic skills used for laparoscopic cholecystectomy and without advanced laparoscopic expertise. MATERIAL AND METHODS: We retrospectively reviewed a 36-month experience (1991 through 1994) at our institution with a consecutive series of patients who underwent basic laparoscopic surgical procedures but specifically excluding cholecystectomy, appendectomy, herniorrhaphy, and colectomy. RESULTS: Procedures performed laparoscopically included gastrostomy, jejunostomy, small bowel resection, intra-abdominal and retroperitoneal biopsy, staging of intra-abdominal malignant lesions, and adhesiolysis for relief of small bowel obstruction. During the 3-year study period, 106 patients underwent 107 procedures, 89 of which were successful. Four patients had substantial complications, two of whom underwent surgical repair. CONCLUSION: A spectrum of procedures may be safely performed with the skills learned from laparoscopic cholecystectomy and without the need for advanced laparoscopic skills such as intracorporeal suturing or tying of knots.

Biopsy↗
PubMed1995

Utilization and outcomes of surgical gastrostomies and jejunostomies in an era of percutaneous endoscopic gastrostomy: a population-based study.

OBJECTIVE: To determine the indications for use of surgical gastrostomy (SG) and surgical jejunostomy (SJ) as feeding tubes, the complications, and the trends in the use of SG and SJ after the introduction of percutaneous endoscopic gastrostomy (PEG) at our institution in 1981. DESIGN: We conducted a retrospective, population-based cohort study of residents of Olmsted County, Minnesota, who received surgically placed feeding tubes between 1976 and 1989. MATERIAL AND METHODS: The medical records of all Olmsted County residents with surgical placement of a feeding tube during the designated study period were reviewed, and underlying conditions, complications, and survival were analyzed. RESULTS: Of 77 adult patients (mean age, 66 years; 48% women), 54 underwent SG and 23 had SJ. General anesthesia was used in 42 patients (55%). The indications for SG or SJ were stroke in 23 patients, cancer in 19, other central nervous system-related conditions in 16, and other conditions in 19. Among the numerous comorbid conditions, pulmonary disease (N = 44) and cardiac disease (N = 32) were most frequent. The median duration of follow-up was 181 days. Complications occurred in 31 of 54 patients (57%) with SG and in 13 of 23 (57%) with SJ. Of the 117 complications, 15% were considered major. Twenty patients (26%) resumed eating. Survival at 1, 6, and 12 months was 79%, 49% and 36%, respectively. Most deaths were due to the disease for which the feeding tube had been placed. In a Cox proportional hazards regression analysis, only age and hypoxemia were found to be significantly associated with survival. Hypoxemia, type of tube, central nervous system disease as indication for procedure, and previous aspiration were associated with failure to resume eating (P < 0.05). Survival was similar to that for our patients with PEG during the same period. The overall incidence of feeding tube placement increased throughout the study period. CONCLUSION: Patients who require enteral feeding tubes have multiple comorbid conditions that have a major influence on the outcome. The overall incidence of feeding tube placement increased after the introduction of PEG. In patients who require long-term enteral nutrition and are unable to have a feeding tube placed percutaneously, surgically placed feeding tubes have outcomes similar to those reported for patients with PEG.

Aged↗
PubMed1995

The surgeon's role in the treatment of chronic intestinal pseudoobstruction.

OBJECTIVE: The surgeon's role in the treatment of patients with chronic intestinal pseudoobstruction (CIP) is under-appreciated. Our aim was to determine the effects of operative treatment on symptomatic relief of CIP. METHODS: Records of all 21 patients who underwent surgery for CIP from 1980 to 1990 were reviewed. CIP was diagnosed by a combination of manometric, radiological, and/or histological examinations. RESULTS: Six of the nine patients who underwent resection or bypass of presumably localized disease are currently maintained on oral intake; one patient with multiple sclerosis who cannot eat is fed via a gastrostomy tube. Enterostomy tube(s) were placed in the other 12 patients; four no longer use the enterostomy tube(s), and eight rely on these decompressive tube enterostomies for symptomatic relief. After operative treatment, hospital readmissions decreased from a mean of 0.5 to 0.1 admission/year. All 19 patients currently alive report improved quality of life. CONCLUSIONS: Selective and directed operative treatment of patients with CIP can result in therapeutic and palliative benefits. The surgeon should have an active role in the evaluation and possible treatment of patients with CIP.

Adolescent↗
PubMed1994

Radical pancreatoduodenectomy for misdiagnosed pancreatic mass.

Occasionally pancreatoduodenectomy is performed for clinically suspected pancreatic malignancy only for the surgeon to find that a benign aetiology accounts for the pancreatic mass. The aim of this study was twofold: to determine the incidence of pancreatoduodenectomy performed for a misdiagnosis of pancreatoduodenal malignancy and to identify potentially avoidable errors in preoperative and intraoperative judgement. Between 1956 and 1990, radical pancreatoduodenectomy was performed in 603 patients at the Mayo Clinic; 29 (5 per cent) underwent pancreatoduodenectomy for a diagnosis made before and during surgery of primary pancreatic or periampullary malignancy that was later proven histopathologically to be either unsuspected subacute or chronic pancreatitis (22 patients), benign fibrous common bile duct stricture (two), and penetrating duodenal ulcer, metastatic melanoma, ampullary adenoma, intrahepatic drug-induced cholestasis and pseudocyst (one each). No patient had a preoperative diagnosis of chronic pancreatitis. Weight loss, pain and/or jaundice were present in 27 of the 29 patients. Objective imaging tests, such as computed tomography, ultrasonography and/or endoscopic retrograde cholangiopancreatography, were performed in 25 patients. Potential errors in judgement may be avoided by a more aggressive attempt at biopsy in selected patients; in others, resection for presumed malignancy may be unavoidable.

Aged↗