Invited Commentary: Zollinger-Ellison Syndrome: A Personal Perspective

By Ellison, E Christopher

The odyssey of Zollinger-Ellison Syndrome is one of the most interesting in surgery. From the initial recognition of two patients with severe peptic ulcer disease in 1955 sprung an international dialogue on this unique disease culminating in discovery that gastrin was the hormone secreted by these pancreatic and duodenal tumors. The impact was the birth of a new area of science, “gastrointestinal endocrinology.” Initially, surgeons were challenged by the complexity of the patients and the need to perform total gastrectomy to prevent death from complications of the severe ulcer disease. Later, after the discovery of proton pump inhibitors, total gastrectomy was no longer needed and the surgeon could focus treatment on tumor removal added by radioimmunoassay for gastrin and new imaging modalities. Today, we recognize that all gastrinomas have malignant potential; early surgical removal can reduce the incidence of metastases and improve survival, independent of biochemical cure, in both sporadic and familial forms of the disease. THIS ARTICLE SUMMARIZES MY personal perspective on Zollinger-Ellison Syndrome. This is based on my initial exposure to the disease from my father, Dr. Edwin H. Ellison, and later from Dr. Robert M. Zollinger, Sr., when I was a resident in surgery and our subsequent experience over the past 50 years at Ohio State University. The objectives of the article are to provide the reader: 1) an overview of the history of the disease description; 2) an algorithm for the evaluation of hypergastrinemia; 3) the best methods of tumor localization; 4) the prognostic factors associated with survival; 5) the presentation of a staging system that will assist in the categorization your patients; and 6) the results of primary tumor resection in this disease. The article is divided into three sections. The first is “Gastrinoma: The Beginning” and includes a review of the first 25 years of the disease with notable historical comments. The second section is entitled “Gastrinoma: The Renaissance,” which is the second 25 years of the disease. The third section is a review of a 50-year experience of gastrinoma at Ohio State University.

Gastrinoma: The Beginning

A relationship between noninsulin-producing islet cell tumors and gastric acid hypersecretion was first advanced by Dr. Robert M. Zollinger, Sr., and Dr. Edwin H. Ellison on April 29, 1955, at the annual meeting in Philadelphia of the American Surgical Association (Fig. 1). The paper was entitled “Primary Peptic Ulcerations of the Jejunum Associated with Islet Cell Tumors of the Pancreas.”1 The article postulated an ulcerogenic humoral factor of pancreatic islet origin and a diagnostic triad for a clinical syndrome was proposed. The diagnostic triad consisted of: 1) the presence of primary ulcerations in unusual locations such as the second or third portions of the duodenum, upper jejunum, or recurrent ulcerations after various gastric procedures; 2) gastric hypersecretion of tremendous proportions; and 3) the identification of non-beta islet cell tumors of the pancreas. The initial cases were both women. One was 36 years old and the other was 19 years old at presentation. The older patient was first seen by the late Edwin H. Ellison. This was on referral on Dr. Harry Burley of Springfield, OH, and Dr. Richard Fulton of Columbus, OH, in July 1952.2 The patient presented with a recurrent massive hemorrhage and underwent multiple resections, subsequently resulting in total gastrectomy and an esophagoduodenostomy on April 14, 1954. The patient died June 7, 1954. A postmortem examination demonstrated no specific tumor of the pancreas, but there was of note hypertrophy and hyperplasia of the adrenal glands and microscopic adenomas of the pancreas, the largest measuring almost 1 cm in diameter within the central portion of the pancreas. The second patient was a 19-year-old single woman who was admitted to the emergency room of the University Hospital on January 19, 1954, a referral from Dr. C. C. Landon of Columbus, OH.2 This patient subsequently underwent a number of operative procedures for recurrent ulceration.

FIG. 1. Robert M Zollinger, Sr., M.D., 1904-1992, and Edwin H. Ellison, 1918-1970.

Indeed, Dr. Zollinger was challenged by this particular case because of the recurrent nature of ulceration after gastric resection. He, therefore, on January 28, 1954, performed a subdiaphragmatic vagotomy, radical gastrectomy with fundusectomy, and end-toend gastroduodenostomy. Of note, the residual gastric pouch measured only 6 x 8 cm. The fundus of the stomach had been removed in an effort to reduce the amount of gastric acid secretion, a procedure recommended in 1931 by Dr. F. G. Connell of Oshkosh, WI. Dr. Zollinger prepared a colored motion picture of the operation because of the unique technical approach being used to control the marked gastric hypersecretion in such a young patient. It was planned to submit this motion picture for review by the Motion Picture Committee of the American College of Surgeons. Within 10 days after the operation, however, the patient developed precordial burning and midline pain. Gastric aspiration at that time showed a 12-hour volume of 570 mL with 28 mEq of free acid. Within 24 hours after discharge from the hospital, she was readmitted because of severe upper abdominal pain and emesis with one episode of bloody emesis.2

Dr. Zollinger wrote to Dr. Hilger P. Jenkins, then Chairman of the Motion Picture Committee of the American College of Surgeons, indicating that our “dream operation” for ulcerative disease had failed and that obviously the motion picture made at the time of fundusectomy, vagotomy, and gastroduodenostomy should not be shown at the Clinical Congress. On August 5, 1954, Dr. Jenkins replied in a letter to Dr. Zollinger “this reminds a bit of a case I had of multiple gastric duodenal and jejunal ulcerations and ultimate perforation in a lady that had an islet cell tumor of the pancreas. She was apparently carrying out a constant insulin test on her gastroduodenojejunal mechanism which was her undoing. Is there any possibility that your little girl as a touch of hyperinsulinism as an etiology for this interesting condition?”3

On receipt of this letter, the patient underwent a 48-hour fast for repeated blood sugar levels to determine if the patient had hypoglycemia; however, there was no hypoglycemia. Subsequently, the patient had recurrent ulcerations that failed various forms of management and, on November 3, 1954, she underwent a total gastrectomy. At that time, a penetrating esophageal ulcer proximal to the cardia was found and, in addition, there was a large ulcer, which had practically replaced the posterior side of the gastroduodenal anastomosis. A Roux-en-Y esophagojejunostomy was performed after completion gastrectomy. The adrenal glands were explored, but no tumor was found. Then, in all likelihood, because of Dr. Jenkins’ suggestions, the pancreas was thoroughly explored before closure for evidence of an islet cell tumor. Two small nodules thought to be lymph nodes were excised from the anterior aspect of the left side of the body of the pancreas. Examination of the small excised gastric pouch showed prominent rugai folds and ulcerations as noted previously. Microscopic examination of the two nodules removed demonstrated pancreatic adenoma consistent with a nonbeta islet cell tumor and a lymph node that had a metastatic islet cell tumor. The original presentation was concluded by a discussion section that included additional cases reported by Dr. Hilger P. Jenkins, Dr. Allan O. Whipple, Dr. Lester Dragstead, Dr. Edgar Poth, and Dr. Carl A. Moyer.

Drs. Zollinger and Ellison concluded in the paper “it seems attractive to speculate that the islet cell tumors in these patients might have created excessive amounts of glucagon or some other substance capable of exciting gastric secretion. It would seem that a heretofore unrecognized islet cell tumor of the pancreas with an ulcerogenic potential must be considered in certain atypical cases of peptic ulcer disease.”4 Of note, in the “Discussion” section of the paper is an observation made by Dr. Lester Dragstead, then of Chicago, IL, and the University of Chicago. Dr. Dragstead made the following comment: “it is quite possible that a study of these pancreatic cells may provide clues to the type of cell in the gastric antrum that produces gastrin. This would be most interesting.”5 This certainly foresaw the potential of gastrin as the hormone involved in Zollinger-Ellison Syndrome. Subsequently, the eponym Zollinger-Ellison Syndrome was coined by Dr. Ben Eiseman at the University Surgeons meeting in Indianapolis February 9 to 11, 1956. Dr. Eiseman said “because of the stimulus of Drs. Zollinger and Ellison we were fortunate to discover a patient with a pancreatic tumor associated with fulminating and resistant peptic ulcerations because they first described a condition and brought it to our attention, it is suggested that this clinical entity be called the Zollinger-Ellison Syndrome.”6(p.170)

The identification of the hormone causing Zollinger-Ellison Syndrome subsequently followed based on the observations of Professor R.A. Gregory and Dr. Hilda J. Tracy of the University of Liverpool, Department of Physiology. Dr. Gregory had visited the Department of Physiology and Dr. Ogden, the Professor and Chair at Ohio State University. During his visit, he and Dr. Zollinger met for a discussion of Zollinger-Ellison Syndrome. He wrote to Dr. Zollinger in May 1960 and indicated “that a few weeks ago a case appeared in Birmingham in which the only pathology found at operation was a small tumor the pancreas which was removed with relief of symptoms.”2 Dr. Gregory and Dr. Tracy were fortunate to obtain a portion of this tumor. From this, histamine-free extracts were made. The tumor extracts were injected into conscious dogs with gastric pouches and in each case was noted to cause a large and prolonged secretory response identical in character with that produced in the same animals by the injection of a gastrin extract made from hog antral mucosa. A gastrin-like activity present in the tumor was the equivalent of 33 to 45 times its weight of hog antral mucosa. Subsequently, the investigators reported the experiments in The Lancet entitled “Extraction of Gastrin-Like Substance from a Pancreatic Tumor in a Case of Zollinger-Ellison Syndrome.”7 Indeed, this supported the role of gastrin in these cases. During the same period of time, Underdahl observed a syndrome of multiple endocrine adenomas.8 Subsequently, Paul Wermer from Columbia University presented a manuscript entitled “Genetic Aspects of Adenomatosis of Endocrine Glands.”9 In this article, Dr. Wermer noted “In a very high percentage of these patients chronic gastric and duodenal ulcers were found. Among five patients who belong to the family here reported there were two instances of gastric ulcer and two of duodenal ulcer. Among the eight cases reported by Underdahl there were three patients with duodenal ulcers and one of these also had a gastric ulcer. . . . This ulcer may represent one more manifestation of the abnormal gene. The high incidence of peptic ulcer suggests that the penetrance of the gene for the formation of ulcer is not different from that for adenomatosis.” Little did he know that his observations and those presented by Zollinger and Ellison would intersect when it became appreciated that approximately 25 per cent of patients with Zollinger-Ellison Syndrome have multiple endocrine neoplasia Type I. As an interesting sidebar, in the early 1990s, Dr. Sam Wells visited Columbus, OH, to assess our patients with Zollinger-Ellison Syndrome for multiple endocrine neoplasia Type I. One of the patients he screened was the first patient with Zollinger- Ellison syndrome. He noted an elevated serum calcium and parathyroid hormone level. She subsequently underwent a reassessment and was found to have hyperparathyroidism. She had a subtotal parathyroidectomy. Subsequently, her daughter was identified to have hypergastrinemia and hyperparathyroidism. This first patient with Zollinger-Ellison syndrome died on March 9, 2001, some 47 years after her initial treatment, after a pulmonary embolism, which occurred after a thoracotomy and right lower lobectomy for a carcinoid tumor on March 5, 2001.

Although initial reports emphasized the pancreatic location of tumors, a report by Dr. Harry Oberhelman in 1961 indicated that tumors associated with the ulcerogenic syndrome could also occur in the duodenum.10 This was important in broadening our understanding of the multifocal nature of this disease.

Another major historical development pertinent to gastrinoma and Zollinger-Ellison Syndrome was the description of the immunoassay of endogenous insulin in humans by Drs. Rosalyn S. Yalow and Solomon A. Berson in 1960.11 This landmark contribution to the understanding of radioimmunoassay as it applies to the measurement of small peptides in serum subsequently was applied to the measurement of gastrin by Dr. James D. McGuigan in 1966. 12 This subsequently would allow the tumors to be detected preoperatively.

Before 1966, we had been seeing approximately two to six cases per 5-year period and, after 1966, this almost doubled and remained at about the same level through the current year, coinciding with the radioimmunoassay of gastrin (Fig. 2). Of the 43 patients seen in this initial era, 25 per cent had multiple endocrine neoplasia I (MEN-I). In the early years, treatment was focused on the gastric hypersecretion as well as the surgical removal of the primary tumor. The only effective treatment of the gastric acid hypersecretion was a total gastrectomy. All other treatments failed. Of the 43 patients treated in this first period of time, 1955 to 1980, 80 per cent required total gastrectomy (Fig. 3). There was an aggressive approach to tumor resection in this early group.

Dr. Robert M. Zollinger, Sr., best summarized the men-current thinking about the surgical management of gastrinoma in a film prepared for the American College of Surgeons in 1972:

“. . . This film is concerned with some of the current concepts in the physical management of the patient with an islet cell tumor of the pancreas of the ulcerogenic type. Total gastrectomy was recommended when mis syndrome was introduced in 1955 and about 800 reported cases support this recommendation in the majority of instances. All too many patients have required multiple gastric procedures with an overall mortality rate of about 33%. This is in contrast to the mortality rate of less than 18% from total gastrectomy performed at the time of the first operation. These tumors continue to challenge surgeons for a number of reasons. In the first place, he has a role to perform a total gastrectomy on a lesion that may be benign and, secondly, it is against his principles to remove a normal organ and leave cancer behind as you must do when you have malignant tumors throughout the pancreas. Thirdly, he is tempted to remove what might appear to be a solitary tumor and avoid total gastrectomy. Lastly, his judgment is tormented when he is unable to find a tumor and wonders: what is the best thing to do in the best interest of the patient.”

FIG. 2. Frequency of diagnosis of gastrinoma at Ohio State University, 1955-2005.

FIG. 3. The frequency of gastric procedures for control of acid secretion in gastrinoma. The decreased frequency after 1980 is the result of the introduction of H2 receptor antagonists and proton pump inhibitors.

Gastrinoma: The Renaissance Period: Second 25 Years

During the second 25 years of the history of gastrinoma, several important discoveries occurred that has impacted our management and diagnosis. First, effective medical acid control was possible. This development of the H2 antagonists allowed effective medical treatment in many but not all patients with gastrinoma. The proton pump inhibitors (PPIs) were then developed and have essentially replaced total gastrectomy as the preferred management for acid hypersecretion. Nearly all patients can have successful treatment of acid hypersecretion with PPIs, but some patients may require very high doses, even four to six times mat recommended for typical acid peptic disease.

The discovery of specific methods to firmly establish the biochemical diagnosis occurred during this period of time as well. This included evaluation of hypergastrinemia as well as diagnosis of multiple endocrine neoplasia Type I. The differential diagnosis of hypergastrinemia includes pernicious anemia, chronic PPI utilization, gastric outlet obstruction, retained antrum, G-cell hyperplasia, and gastrinoma. These conditions can be separated from gastrinoma by a secretin provocative test. It was recognized and published by Isenberg et al. in 1972 that there was a paradoxic effect of secretin on serum gastrin in patients with suspected Zollinger-Ellison syndrome.13 Figure 4 demonstrates the effect of intravenous administration of secretin 2 [mu]g/kg in patients with peptic ulcer disease and chronic PPI, pernicious anemia, ZollingerEllison Syndrome, and Zollinger-Ellison Syndrome and PPI. As Tables 1 and 2 demonstrate, after administration of secretin in patients with hypergastrinemia from nontumor-related causes, the serum gastrin level remains flat, whereas in patients with ZollingerEllison Syndrome, whether on PPIs or off PPIs, there is a paradoxic response and increase in serum gastrin. Serum gastrin increase of greater than 110 [mu]g/mL is considered diagnostic, but most often the difference is higher than that. An important point of the provocative testing is that patients with Zollinger-Ellison Syndrome do not need to have the PPI stopped for the secretin provocative test. In other words, the paradoxic response is maintained. The accuracy of this test in my experience is 100 per cent. I have not seen falsepositive or -negative tests, although the National Institutes of Health reports an overall accuracy of 97 per cent.14

Very importantly, the screening for MEN-1 should be done in all patients. This should include measurement of prolactin, calcium, parathyroid hormone, and pancreatic polypeptide, which may be a marker for this condition in some patients. The diagnosis of hyperparathyroidism may occur at variable periods of time during the natural history of MEN-1 associated with Zollinger-Ellison Syndrome. In our experience, approximately one-half of the patients initially present with hyperparathyroidism before the diagnosis of gastrinoma and half of the patients present with the diagnosis of hyperparathyroidism after the establishment of hypergastrinemia and the diagnosis of ZollingerEllison Syndrome. Although in the majority of patients, the diagnosis of gastrinoma and hyperparathyroidism occurs within 10 years of each other, the clinician needs to be aware that late expression may occur (Fig. 5). Indeed, the first patient with ZollingerEllison syndrome was diagnosed nearly 40 years after her original operation with hypercalcemia and hyperparathyroidism.15

FIG. 4. Secretin stimulation in various conditions causing hypergastrinemia. It is not necessary to discontinue proton pump inhibitors before the test because the paradoxic response in gastrinoma is preserved when the patient remains on the proton pump inhibitor. TABLE 1. Elements for Staging Gastrinoma

TABLE 2. Staging of Gastrinoma in 106 Patients

New techniques in imaging have evolved and are extremely important in the management of ZollingerEllison Syndrome today. Angiography, although quite helpful early on in the late 1970s to early 1980s, has been replaced with imaging with CT and MRI. In our experience, MRI is more accurate for detecting metastatic disease. The role of somatostatin scintigraphy has been well outlined in the literature.16 It is extremely valuable in detecting both primary and metastatic disease. In particular, patients with unsuspected metastatic disease can be identified by using octreotide scan. Endoscopic ultrasound has been a valuable test for localization for insulinomas and other endocrine tumors of the pancreas. Although it has an accuracy of nearly 95 per cent for the detection of insulinomas, the technique is approximately 70 per cent sensitive in detection of gastrinomas (Gerry Dougherty, University of Michigan, personal communication, Oct. 10, 2007). Endoscopic ultrasound can visualize tumors in the pancreas, duodenum, and lymph nodes and provide access for biopsy of the lesions.17-19

FIG. 5. Timing of diagnosis of gastrinoma and hyperparathyroidism in multiple endocrine neoplasia I.

An extremely helpful test in patients in whom standard imaging fails to demonstrate a tumor is the selective arterial secretin test described by Imamura.20 In this test, a catheter is placed through the femoral vein into the right hepatic vein. The patient then undergoes selective cannulation of the splenic, hepatic, gastroduodenal, and then the superior mesenteric arteries. Small doses of secretin are selectively injected into the circulation as defined previously. Blood is sampled from the right hepatic vein at baseline and 20, 40, 60, 90, and 120 seconds for gastrin determination. The test has been extremely helpful in cases in which there is no primary tumor identified on other testing.

Figure 6 is an example of a selected arterial secretin test done on a patient who was referred with recurrent hypergastrinemia after excision of what appeared to be a lymph node primary 4 years before his evaluation at Ohio State University. This patient had undergone endoscopic ultrasound that demonstrated a nodule close to the neck of the pancreas. Other imaging was negative with somatostatin scintigraphy included. Biopsy of the lesion was inconclusive. The patient had selective arterial secretin performed that showed a step- up in the distribution of the superior mesenteric artery with values going from 1500 baseline to 4500. Elevation of distribution of the superior mesenteric artery (SMA) would indicate either tumor involving the uncinate process or perhaps the third or fourth portion of the duodenum. At exploration, a small lymph node was identified superior to the neck of the pancreas and it was removed. This was a reactive lymph node and was compatible with the findings on endoscopic ultrasound. Subsequently, the duodenum was opened and examined carefully. This demonstrated a lesion at the junction of the third and fourth portion of the duodenum, which was locally excised. Postoperatively, the patient’s serum gastrin levels returned to normal, and it has remained normal since this operation.

FIG. 6. Example of a selective arterial secretin stimulation test.

Collective analysis indicates that the overall accuracy of imaging tests in localizing gastrinoma is approximately 70 per cent. In 30 per cent of patients, localization fails to indicate the site of the gastrinoma. In these cases, thorough exploration of the pancreas and duodenum is necessary to localize the primary tumor.

In this regard, two major contributions made during this period have further defined the surgeon’s operative approach in this disease. First was the recognition of the gastrinoma triangle by Drs. Bruce Stabile and Ed Passaro.21 This important contribution noted that approximately two-thirds of all gastrinomas are located within the gastrinoma triangle as shown in Figure 7. Our experience has paralleled that of the Los Angeles group. Finally, Dr. Norm Thompson at the University of Michigan and Dr. Jeff Norton at Stanford University have indicated the importance of a duodenotomy in the management of Zollinger-Ellison Syndrome.22 Indeed, one cannot depend alone on imaging tests, transduodenal palpation, or transduodenal illumination with endoscopy to identify duodenal adenomas. A duodenotomy must be made to find these tumors. More often, in sporadic cases, they are the first or second portion of the duodenum, although, as indicated in the case presented here, they can occur in other areas and the surgeon must be diligent in searching the third and fourth portions of the duodenum. Duodenotomy should be considered a mandatory part of exploration for gastrinoma.

FIG. 7. The gastrinoma triangle.

Figure 8 demonstrates a current algorithm for the diagnosis of suspected gastrinoma in 2008. Today, the biochemical diagnosis of gastrinoma should be made by establishing an elevated fasting gastrin off PPIs and verifying with a positive secretin provocative test. Although ideally, gastric secretions should be sampled for acid determination, this is infrequently done because of unavailability of acid titration in most hospital laboratories. Indeed, hypergastrinemia and achlorhydria off H2 antagonists and PPIs for 72 hours is essentially diagnostic of pernicious anemia and atrophic gastritis as a cause of the hypergastrinemia. Effective medical control of acid secretion is best accomplished with PPIs, although dosing will need to be greater than what is standard for ulcer disease. All patients should be screened for multiple endocrine neoplasia Type I with measurement of calcium, parathyroid hormone, prolactin, and pancreatic polypeptide. Imaging to localize gastrinoma should include CT, MRI, endoscopic ultrasound, somatostatin scintigraphy, and possibly, selective arterial secretin stimulation. I prefer to have two positive tests before exploration of the patient.

With a combined accuracy of 70 per cent, there remains debate about whether the 30 per cent of patients with negative localization tests should be explored. Most “gastrinomologists” today recommend that patients should be explored regardless of the localization test results. This is largely based on the observations of the National Institutes of Health group and our own experience that indicates the major cause of death today in gastrinoma is tumor progression and that metastases infrequently develop when tumors less than 2 cm are removed, whereas larger tumors are more prone to develop metastatic disease.23 Hence, the rational for early exploration is to improve survival by surgical excision of tumors before they reach the size that harkens metastases. Interestingly, review of the Ohio State experience has indicated that patients who have Stage 0 disease, that is, no tumor identified at operation or no tumor identified on radiologic imaging tests, have the best survival of all groups and perhaps one could avoid unnecessary explorations in these patients. Additional information is needed to resolve this controversy. Currently, most authorities would recommend exploration for patients with biochemical Zollinger-Ellison Syndrome in the absence of positive localization tests.

FIG. 8. Algorithm for the evaluation of a patient with hypergastrinemia.

The goal of surgical exploration should be to remove all gross tumor. In my experience, a midline laparotomy provides excellent exposure. A wide Kocher maneuver is performed. The lesser sac should be entered. The pancreas should be completely exposed from the head to the tail. Care must be taken to examine lymph nodes in the area of the porta hepatis and the gastrinoma triangle. Indeed, reports of lymph node primaries have been noted in the past, and this area needs to be carefully dissected to identify these lymph node primaries.20 The pancreas should be examined manually and intraoperative ultrasound should be used to assist identification of small pancreatic tumors. Pancreatic tumors should be enucleated if possible. Larger tumors may require either distal pancreatectomy or pancreaticoduodenectomy. Blind pancreatectomy should not be performed in sporadic gastrinoma. A lateral duodenotomy is made in all cases. The duodenum is inspected by manual palpation through the duodenotomy.24, 25 Duodenotomy may be closed in a single layer longitudinally. The extent of resection in patients with MEN-1 is controversial and ranges from a selective approach with local excision of pancreatic tumors combined with enucleation of duodenal tumors and possible blind distal pancreatectomy to pancreaticodudenectomy.15, 26-29 In general, there is no need to perform acid-reducing procedures in the typical sporadic or MEN-1 patient, although in certain cases, highly selective vagotomy or vagotomy and pyloroplasty may be warranted. Acid control should be achieved with PPIs. Total gastrectomy today is rarely if ever indicated. Indeed, in our experience with 63 patients diagnosed since 1980, 70 per cent of the patients required no gastric procedure, 15 per cent a lesser gastric operation, and only 10 per cent required total gastrectomy. All patients requiring total gastrectomy were early on in this experience, predating PPIs, and were cases with complications of severe recurrent ulcer disease such as gastrojejunal colic fistula or being refractory to medical management with H2 antagonists. Experience with laparoscopic approaches is increasing and appears promising.30

Current Controversies and Conclusions: The Ohio State University Experience With Gastrinoma

The major controversies today include, but are not limited to, staging, resection results, and whether patients with MEN should undergo resection. In 2006, we reported an analysis of 106 patients treated at the Ohio State University.31 The minimum period of postdiagnosis observation was 6 years or until death, and more than 90 per cent of patients had a minimum 10-year follow up. The patients were defined as sporadic gastrinoma (nonfamilial) or gastrinoma associated with MEN. They were further categorized using the TNM elements in Tables 1 and 2 as reported in 1995.32 Resection of the gastrinoma was defined as follows: complete removal of the tumor with negative margins and normal postoperative fasting gastrin (R0); resection with microscopic margins and/or elevated postoperative fasting gastrin (Rl); resection with gross residual disease (R2); tumor not resected (NR); and no tumor identified (NT). End points included diseasefree survival (DFS), defined as survival with normal fasting gastrin and imaging negative for recurrence, and disease-specific survival (DSS). We found that staging according to this method was highly reliable in predicting DSS. Interestingly, lymph node involvement was not identified as a predictor of survival or disease progression. The most important variables in staging were tumor size and the presence or absence of liver or other sites of distant metastases. Tumor size rather than location in the duodenum or pancreas was an important predictor of outcome. In addition, we identified that the impact of resection on survival was significant. R0/R1 resection tended to improve survival in all stages and significantly so in Stage 2 and Stage 3. Staging influenced the resection outcome of surgery. In Stage 1,18 per cent of surgical procedures resulted in R2 compared with 41 per cent in Stage 2 and 81 per cent in Stage 3 (P = 0.0001). Twenty-two patients, 21 per cent of all 106 study patients and 30 per cent of 72 patients undergoing resection, had RO resection accomplished at surgery. The median DFS was 12.5 years (range, 1-39.2 years) and was independent of stage. Twelve of these patients are alive and disease-free. Two patients died with no evidence of recurrent gastrinoma. Eight patients (36%) developed recurrence. The median time to recurrence was 8 years. Liver metastases occurred in five an average of 7.2 years (range, 3-13 years) after surgery. All of these patients received additional treatment. Three patients had subsequent hepatic resections and the other two had adjuvant therapy. Of the three having hepatic resection, one died of recurrent disease at 6.4 years, one is alive with recurrent disease at 3.5 years and has received subsequent chemoembolization, and one has no evidence of recurrent disease at 10 years. Three patients had recurrent hypergastrinemia at 1, 7, and 10 years with negative imaging tests. One died in an accident, one of cirrhosis, and one of nutritional complications of total gastrectomy.

Regarding sporadic and MEN patients, we found no difference in survival between these groups. With resection in 22 patients with MEN, we observed postoperative normal gastrin in only one patient who remains eugastrinemic. The incidence of cure rates found in most series of patients with MEN is 5 per cent to 10 per cent with the exception of the experience at the University of Michigan in which cure rate over 30 per cent has been reported. However, in patients with MEN with R0/R1, resection resulted in an 80 per cent survival at 20 years compared with 40 per cent in those who did not have resection. Hence, regardless of the inability to cure patients with MEN, tumor excision confers a survival advantage perhaps by reducing eventual metastases.

In summary, Zollinger-Ellison syndrome is of historic significance in gastrointestinal surgery. It is a story of two surgeons, two patients, and the dialogue of international surgery and physiology of the mid20th century. The Odyssey over the past 50 years has resulted in a clear definition of the surgical principles in the treatment of this challenging disease. The incidence has remained constant over the past number of years. There will be approximately one case per million population per year in most areas. Gastrinoma should be categorized as sporadic or associated with multiple endocrine neoplasia. All lesions should be considered malignant and staged according to a TNM system with the caveat that lymph node metastases do not negatively affect survival. The diagnosis is confirmed by hypergastrinemia and the positive secretin test. Localization tests should be performed in all patients before operative intervention. Surgery should consist of careful exploration of the pancreas and duodenum for tumor excision. The goal is an R0/R1 resection. Enucleation of tumors is preferred, although larger tumors may require a pancreatic resection. Survival rates for resected tumors should exceed 10 years in over 80 per cent of patients. Patients with isolated tumors and MEN-I should undergo resection as well even though cure is unlikely; survival benefit is noted in this group of patients. Survival in both sporadic and MEN cases is dependent on the extent of resection and stage but is independent of achieving a normal postoperative gastrin level. After treatment, it is recommended that patients have an annual fasting gastrin and imaging based on gastrin levels and symptoms. Challenges for the future include genetic screening for multiple endocrine neoplasia Type I, further definition of the extent of resection in multiple endocrine neoplasia Type I, the possibility of percutaneous or endoluminal treatment for small primary tumors, perhaps by endoscopic ultrasonography-directed therapies, establishing practice guidelines for the treatment of metastatic disease with resection and/or ablative techniques or hepatic artery chemoembolization, and, finally, the potential use and experimental modality of radiolabeled somatostatin for treatment of metastatic disease.

REFERENCES

1. Zollinger RM, Ellison EH. Ulcerations of the jejunum associated with islet cell tumors of the pancreas. Ann Surg 1955; 142: 709-28.

2. Zollinger RM, Coleman DW. The Influences of Pancreatic Tumors on the Stomach. Springfield, IL: Charles C. Thomas, 1974.

3. Zollinger RM, Coleman DW. The Influences of Pancreatic Tumors on the Stomach. Springfield, IL: Charles C. Thomas, 1974:12.

4. Zollinger RM, Ellison EH. Ulcerations of the jejunum associated with islet cell tumors of the pancreas. Ann Surg 1955; 142: 722.

5. Zollinger RM, Ellison EH. Ulcerations of the jejunum associated with islet cell tumors of the pancreas. Ann Surg 1955; 142: 724.

6. Ellison EH. The ulcerogenic tumor of the pancreas. Surgery 1956;40:147-70.

7. Gregory RA, Tracy HJ, French JM, et al. Extraction of a gastrin-like substance from a pancreatic tumor in a case of Zollinger-Ellison syndrome. Lancet 1960; 1:1045.

8. Underdahl LO, Woolner, Black BM. Report of 8 cases in which the parathyroids, pituitary, and pancreatic islets were involved. J Clin Endocrinol 1953;13:20.

9. Wermer P. Genetic aspects of adenomatosis of the endocrine glands. Am J Med 1954;16:363.

10. Oberhelman HA, Nelson TS. Surgical considerations in the management of ulcerogenic tumors of the pancreas and duodenum. Am J Surg 1964;108:132.

11. Berson SA, Yalow RL. Immunoassay of endogenous insulin in man. J Clin Invest 1960;39:1157-75.

12. McGuigan JD, Trudeau WL. Immunochemical measurement of elevated levels of gastrin in the serum of patients with pancreatic tumors of the Zollinger-Ellison variety. N Engl J Med 1966;298:1308- 15.

13. Isenberg JI, Walsh JH, Passare E, et al. Unusual effect of secretin on serum gastrin, serum calcium, and gastric acid secretion in a patient with suspected Zollinger-Ellison syndrome. Gastroenterology 1972;62:626-31.

14. Berna MJ, Hoffman KM, Long SH. Serum gastrin in Zollinger- Ellison syndrome: II. Prospective study of gastrin provocative testing in 293 patients from the National Institutes of Health and comparison with 537 cases from the literature. Evaluation of diagnostic criteria, proposal of new criteria, and correlations with clinical and tumoral features. Medicine (Baltimore) 2006;85:331-64.

15. Melvin WS, Johnson JA, Sparks J, et al. Long-term prognosis of Zollinger-Ellison syndrome in multiple endocrine neoplasia. Surgery 1993;114:1183-9.

16. Ellison EC, Schirmer WJ, Olsen JO, et al. Localization of neuroendocrine tumors using somatostatin receptor imaging with indium-lll-pentetreotide (OctreoScan). Cancer Control 1997;4: 35-9.

17. Noone TC, Hosey J, Firat Z, et al. Imaging and localization of islet-cell tumours of the pancreas on CT and MRI. Best Pract Res Clin Endocrinol Metab 2005;19:195-211.

18. McLean AM, Fairclough PD. Endoscopic ultrasound in the localization of pancreatic islet cell tumours. Best Pract Res Clin Endocrinol Metab 2005;19:100-14.

19. Gauger PG, Scheiman JM, Wamsteker EJ. Role of endoscopic ultrasonography in screening and treatment of pancreatic endocrine tumours in asymptomatic patients with multiple endocrine neoplasia type 1. Br J Surg 2003;90:748-54.

20. Imamura M, Takahashi K. Use of selective arterial secretin injection test to guide surgery in patients with Zollinger-Ellison syndrome. World J Surg 1993;17:433-8.

21. Stabile BE, Morrow DJ, Passare E Jr. The gastrinoma triangle: operative implications. Am J Surg 1984;147:25-31.

22. Norton JA, Alexander HR, Fraker DL, et al. Does the use of routine duodenotomy affect rate of cure, development of liver metastases, or survival in patients with Zollinger-Ellison syndrome? Ann Surg 2004;239:617-25.

23. Norton JA, Fraker DL, Alexander HR, et al. Surgery increases survival in patients with gastrinoma. Ann Surg 2006;244: 410-9.

24. Norton JA, Alexander HR, Fraker DL, et al. Possible primary lymph node gastrinoma: occurrence, natural history, and predictive factors: a prospective study. Ann Surg 2003;237:650-7.

25. Zogakis TG, Gibril F, Libutti SK, et al. Management and outcome of patients with SG arising in the duodenum. Ann Surg 2003;238:42-8. 26. Thompson NW. Current concepts in the surgical management of multiple endocrine neoplasia type 1 pancreatic- duodenal disease. Results in the treatment of 40 patients with ZollingerEllison syndrome, hypoglycemia or both. J Intern Med 1998;243: 495-500.

27. Norton JA, Fang TD, Jensen RT. Surgery for gastrinoma and insulinoma in multiple endocrine neoplasia type 1. J Natl Compr Canc Netw 2006;4:148-53.

28. Bartsch DK, Fendrich V, Langer P, et al. Outcome of duodenopancreatic resections in patients with multiple endocrine neoplasia type 1. Ann Surg 2005;242:757-64.

29. Norton JA, Jensen RT. Resolved and unresolved controversies in the surgical management of patients with Zollinger-Ellison syndrome. Ann Surg 2004;240:757-73.

30. Assalia A, Gagner M. Laparoscopic pancreatic surgery for islet cell tumors of the pancreas. World J Surg 2004;28:1239-47.

31. Ellison EC, Sparks J, Verducci JS, et al. 50-year appraisal of gastrinoma: recommendations for staging and treatment. J Am Coll Surg 2006;202:897-905.

32. Ellison EC. Forty-year appraisal of gastrinoma: Back to the future. Ann Surg 1995;222:511-24.

E. CHRISTOPHER ELLISON, M.D.

From the Department of Surgery, Ohio State University Medical Center, Columbus, Ohio.

Editor’s Note: Dr. Ellison is currently Chair of the Department of Surgery at Ohio State University Medical Center, which was the position occupied for many years by Robert Zollinger. As noted in his introduction, he is the son of Dr. Edwin Ellison, the other surgeon who made the sentinel observations on Zollinger-Ellison syndrome.

Address correspondence and reprint requests to E. Christopher Ellison, M.D., 327 Means Hall, 1652 Upham Drive, Columbus, OH 43210. E-mail: [email protected].

Copyright Southeastern Surgical Congress Jul 2008

(c) 2008 American Surgeon, The. Provided by ProQuest LLC. All rights Reserved.