Prospective study of gastrinoma localization and resection in patients with Zollinger-Ellison syndrome

Current diagnosis and management of Zollinger-Ellison syndrome

The treatment of patients with Zollinger-Ellison syndrome (ZES) has undergone dramatic evolution during the past decade. Although initially regarded as an incurable tumor, resection of gastrinoma for potential cure has been reported in 30% to 40% of selected patients in recent series. Conversely, although definitive control of acid hypersecretion is achieved by total gastrectomy, histamine (H2)-receptor antagonists and the newly introduced agents omeprazole and somatostatin analogues allow effective medical therapy of gastric acid overproduction. Confirmation of the diagnosis is best achieved with the I.V. secretin stimulation test, and tumor localization techniques are mandatory to identify candidates for operative tumor resection. Intraoperative sonography and careful exploration are required for tumor removal; successful tumor resection is associated with prolonged survival. The majority of patients (60%) are still found to have malignant disease at the time of diagnosis, but 10-year overall survival commonly exceeds 40%. The presence of multiple endocrine neoplasia type I (MEN-I) is seen in 10% to 25% of patients; correction of hypercalcemia alone may have therapeutic benefit in some ZES patients, and while gastrinoma resection is rarely possible, MEN-I patients demonstrate prolonged survival. The choice of medical rather than surgical therapy for acid hypersecretion depends on the suitability of each patient for careful and repeated endoscopic and chemical studies, versus the likelihood of a successful postoperative outcome. Socioeconomic, geographic, and related medical factors in each case may dictate the form of long-term antisecretory therapy. Exploration for possible tumor resection is indicated for virtually all patients who have no documented metastatic disease.

Curative resection of multiple gastrinomas aided by selective arterial secretin injection test and intraoperative secretin test

Recently a number of surgeons have recommended radical resection of gastrinomas in Zollinger-Ellison syndrome (ZES). We have developed a useful technique for preoperative localization of gastrinomas--the selective arterial secretin injection test (SASI)--and we recommend an intraoperative secretin test (IOS) for deciding the radicality of resection of gastrinomas. Here the results of SASI and IOS tests in 11 patients with ZES are examined and compared with the results of other techniques. The SASI test localized gastrinomas in all of the patients, while the sensitivity of ultrasonography, computed tomography, arteriography, or portal venous blood samplings was between 1/11 and 5/11. On the basis of the results of the SASI test, radical resection of gastrinoma was performed in four patients (three pancreatoduodenectomies and one extirpation). After pancreatoduodenectomy, immunohistologic study of the specimen revealed multiple microgastrinomas and lymph node metastases in two patients and the coexistence of a microgastrinoma and a gastinoma in one patient. The IOS test was useful in the estimation of the advisability of radicality, and in two patients total gastrectomy was not performed because of the results of the IOS test. These four patients are well and have returned to work, and their serum gastrin levels are below 35 pg/mL. Thus we believe SASI and IOS tests are helpful for planning curative resection of gastrinomas.

Zollinger-Ellison Syndrome in the Intensive Care Setting

Management of patients with Zollinger-Ellison syn drome differs markedly from management of patients with idiopathic gastric acid hypersecretion or routine peptic ulcer disease. Because of the possible complica tions arising from gastric acid hypersecretion or from complications caused by the gastrinoma itself, patients with these disorders frequently present to critical care physicians. It is important that critical care physicians be familiar with the acute presentation of patients with Zollinger-Ellison syndrome as well as the appropriate treatment. We review the important points pertaining to the recognition and treatment of Zollinger-Ellison syn drome in the acute care setting.

Gastrin-producing ovarian cystadenocarcinoma: sensitivity to secretin and SMS 201-995

We report a patient with severe peptic ulcer disease and a right ovarian mass that was found to be a gastrin-producing cystadenocarcinoma. Gastrin production by the tumor was stimulated by secretin and inhibited by the long-acting somatostatin analogue SMS 201-995. Following resection of the tumor, serum gastrin levels and the gastrin response to secretin returned to normal. Histologic examination, including Alcian blue staining for mucin and immunoperoxidase staining for gastrin, revealed gastrin at the base and mucin at the apex of the tumor cells. This report demonstrates secretin stimulation and somatostatin inhibition of gastrin secretion from a cell that is apparently not of endocrine origin.

Ovarian carcinoma as a cause of Zollinger-Ellison syndrome. Natural history, secretory products, and response to provocative tests

Zollinger-Ellison syndrome is usually caused by a gastrin-secreting tumor in or near the pancreas. We describe a patient in whom an ovarian cystadenocarcinoma was the cause of the syndrome. The patient presented with a short history of peptic ulceration and development of a large pelvic mass. Investigations demonstrated a basal acid output of 37.8 mEq/h and a maximal acid output of 36.0 mEq/h, and the plasma concentration of gastrin was 830 pg/ml (normal less than 100). Secretin and calcium infusion tests were positive, and a meal test was compatible with Zollinger-Ellison syndrome. Imaging studies demonstrated a normal liver and pancreas but a large cystic right ovarian mass. Resection of the mass resulted in a marked reduction in gastric acid output, a fall in plasma gastrin concentration to normal, negative calcium and secretin tests, and a normal (positive) meal test. Histology of the mass showed it to be a mucinous cystadenocarcinoma. The tumor stained with immunoperoxidase technique was positive for gastrin, and the cyst fluid contained high concentrations of gastrin and calcitonin. One year later, the patient has no biochemical or imaging evidence of tumor. Ovarian, gastrin-producing tumors and pancreatic gastrinomas cannot be distinguished by provocative tests, and negative imaging studies do not exclude a pancreatic tumor. Patients with an ovarian mass and Zollinger-Ellison syndrome should have a bilateral oophorectomy and a careful exploration of the pancreatic area.

Parenteral antisecretory drug therapy in patients with Zollinger-Ellison syndrome

Forty-six patients with Zollinger-Ellison syndrome were studied prospectively to determine a safe and effective method and criterion for controlling gastric acid hypersecretion during periods when oral antisecretory agents could not be used. In each patient it was possible to reduce acid secretion to less than or equal to 10 mEq/h after an i.v. bolus of 150 or 300 mg of cimetidine and a stepwise titration of cimetidine given by continuous infusion. The mean dose given by i.v. infusion was 2.9 mg/kg body wt.h but there was a wide range (0.5-7.0 mg/kg body wt.h) and the minimal dose had to be determined individually for each patient. The minimal i.v. cimetidine dose did not correlate with basal or maximal acid output or fasting gastrin concentration, but correlated closely with either the previous oral dose of cimetidine (r = 0.96, p less than 0.001) or the previous oral dose of ranitidine or famotidine (r = 0.95, p less than 0.001). To study the efficacy and safety of an i.v. infusion of cimetidine, 34 patients undergoing surgery were maintained on i.v. cimetidine for a mean of 12 days (range 1-83 days). One-half of the patients did not require dose adjustment, whereas the remainder required an average of 2 adjustments, usually in the first 3 postoperative days. No patient developed complications attributable to gastric acid hypersecretion in the postoperative period, and there was no detectable neurologic, hematologic, or hepatic toxicity. This study demonstrates that a continuous i.v. infusion of cimetidine adequately inhibits gastric acid hypersecretion in patients with Zollinger-Ellison syndrome. However, high doses were frequently required, the dose had to be determined in a stepwise fashion individually for each patient, and the i.v. dose correlated with the previous oral dose. Reducing acid secretion to less than or equal to 10 mEq/h was a safe criterion during surgery and continuous i.v. cimetidine was safe and effective in achieving this degree of control for up to 83 days.

Gastrinoma. New medical and surgical approaches

The clinical presentation of gastrinoma has changed since the original description by Zollinger and Ellison in 1955. Tumors currently found are smaller, extrapancreatic or extraintestinal in location, and frequently occult within lymph nodes. The incidence of hepatic metastases on initial presentation has decreased. In addition, the clinical course of patients with tumor in lymph nodes is benign, suggesting that more patients than were previously thought are now candidates for cure. Improved knowledge of the anatomic location of gastrinomas has enhanced our ability to find and remove them at laparotomy. As a result of these factors, more patients are being cured than ever before, and in the future, cure rate may be even higher. On the basis of these recent advances, the optimal treatment of gastrinoma is surgical excision for cure.

Microgastrinomas of the duodenum. A cause of failed operations for the Zollinger-Ellison syndrome

Gastrinomas are now being detected at an earlier stage than was formerly the case. Furthermore, with the ability to control acid secretion, emphasis has been placed on identifying gastrinoma patients who are potentially curable by tumor resection rather than by palliative gastrectomy. Despites estimates suggesting that 20-40% of sporadic gastrinoma patients can be successfully resected for cure, as many as 40% of such patients have occult tumors that elude detection. In an effort to better localize gastrinomas, we have used percutaneous transhepatic venous (THVS) gastrin sampling over the past 10 years. From 1978 to 1988, THVS was used in 46 patients in whom there was no other evidence of metastatic gastrinoma by conventional studies. Gastrinomas were found at operation in all but one patient. The purpose of this report is to emphasize that occult tumors are most often found in the duodenal wall, and frequently they may be no greater than 2 mm in diameter. Five recent cases illustrate that these small tumors or microgastrinomas may be the sole source of hypergastrinemia and can be cured by local excision. These recent cases emphasize that microgastrinomas are not usually palpable through the duodenal wall. They may be detected only after duodenotomy and meticulous evaluation of the mucosa by eversion and direct palpation. Duodenotomy and intraluminal exploration should be considered an essential component of the operation for patients with extrapancreatic gastrinomas.

Use of long-acting somatostatin analog SMS 201-995 in patients with pancreatic islet cell tumors

Natural somatostatin reduces plasma concentrations of many peptides, and is of short term benefit in patients with islet cell tumors, but has to be given as a continuous intravenous infusion. We review the published experience with the long acting synthetic somatostatin analogue SMS 201-995 in patients with islet cell tumors. Fifteen of 18 patients with vasoactive intestinal peptide-producing tumors, 8 of 8 patients with glucagonomas, 7 of 7 patients with unresectable insulinomas, and 3 of 3 patients with growth hormone releasing factor-producing tumors had a good sustained symptomatic response to SMS 201-995. Patients with benign insulinomas responded variably and are best treated by surgery. Patients with gastrinomas are best treated by oral gastric antisecretory agents. In all these syndromes, the clinical response to SMS 201-995 did not necessarily parallel the change in plasma concentration of marker peptide, suggesting that SMS 201-995 may have actions at various sites. The effect of SMS 201-995 on tumor size has been assessed in 46 patients, less than 20% of whom showed a reduction in tumor size. Side effects have been mild, but include steatorrhea and gastrointestinal disturbances. More studies will be required to fully assess the effects of long-term administration of SMS 201-995.

Effects of omeprazole on acid secretion and acid-related symptoms in patients with Zollinger-Ellison syndrome

In the Zollinger-Ellison syndrome, symptoms and complications are due to hypersecretion of acid, and the first therapeutic step is to suppress the acid secretion. Long-term treatment with histamine H2-receptor antagonists was compared with omeprazole treatment. A total of 30 consecutive ZES patients were treated continuously with H2-receptor antagonists. During long-term treatment, a marked tachyphylaxis was noted, more than 50% of the patients had periods of dyspepsia, recurrent ulcers were found in 10 patients and in 16 a decline in the action of the H2-receptor antagonist required a change to omeprazole after a median duration of 36 months. A total of 22 patients were treated with omeprazole. During long-term treatment, the dose could be reduced slightly. Inhibition of acid secretion was maintained in all cases, and none had dyspeptic symptoms. The median duration of treatment was 18 months, with a range of 1-120 months (H2-receptor antagonists) and 27 months with a range of 1-66 months (omeprazole). No side-effects were seen with omeprazole.

Surgical treatment of carcinoids and endocrine pancreatic tumours

The common association of neuroendocrine abdominal neoplasms, carcinoids and endocrine pancreatic tumours, with often severe endocrine symptoms has justified considerable efforts to improve treatment in these conditions. Surgery still constitutes the principle therapy for the majority of these tumours. However, the introduction of new means of medical treatment with cytostatic agents, or more recently interferon and a somatostatin analogue, seem to have impact on indications and the extent of surgery in malignant forms of these tumours. It is thus probable that a surgical tumour reduction will increase the possibilities to achieve positive effects of the medical treatment even in advanced malignancies. However, the existence of an alternative medical therapy has increased obligations that surgery should be performed without morbidity or mortality.

Intraoperative ultrasonographic localization of islet cell tumors. A prospective comparison to palpation

The purpose of the present study was to evaluate prospectively the value of intraoperative ultrasound scanning (IOUS) in localizing islet cell tumors by comparing results of IOUS to those of palpation during 44 consecutive laparotomies for gastrinoma (36) or insulinoma (8). All patients had preoperative radiographic imaging studies and selective venous sampling for hormones, which guided the subsequent laparotomy. Any suspicious finding by palpation and/or IOUS was resected. Pathologic evidence of islet cell neoplasm served as the reference standard. Five patients were excluded from analysis because neither palpation nor IOUS had suspicious findings and no islet cell tumor was found. Seven pancreatic insulinomas were found in seven patients. IOUS was as sensitive as palpation at localizing insulinomas. Twenty-three pancreatic gastrinomas were found in 19 patients. IOUS was equal to palpation in the ability to localize gastrinomas. Gastrinomas that were successfully imaged by IOUS were significantly larger than gastrinomas that were not imaged. Twelve extrapancreatic gastrinomas were found in nine patients, and palpation was more sensitive than IOUS at localizing these small duodenal wall tumors. Five patients (11%) had their surgical management changed by IOUS. Two patients had pancreatic tumors (one gastrinoma and insulinoma) enucleated that would not have been found without IOUS, and three patients had resections of pathologically proven malignant islet cell tumors based on sonographic findings. All five patients were cured with short follow-up. The present results demonstrate that palpation and IOUS are complementary because IOUS can image tumors that are not palpable and IOUS can provide additional information concerning malignant potential not detected by palpation.

Medical management of patients with Zollinger-Ellison syndrome who have had previous gastric surgery: a prospective study

We examined prospectively the criteria for medical management in 16 patients with Zollinger-Ellison syndrome who had had previous gastric surgery. Each patient received sufficient antisecretory medication to lower gastric acid output to less than 10 mEq/h during the last hour before the next dose of drug. The 7 patients with a vagotomy but no gastric resection were symptom-free and had no mucosal disease. Of 9 patients with a partial gastrectomy, 7 had mucosal disease, with or without symptoms, and 6 of the 7 patients had acid outputs of 5-10 mEq/h. In these patients, antisecretory medication was increased to reduce output to less than 5 mEq/h and symptoms and mucosal abnormalities resolved in each patient. Patients with Zollinger-Ellison syndrome and a vagotomy can be treated safely by reducing acid secretion to less than 10 mEq/h, but in patients with a partial gastrectomy, acid secretion must be reduced to less than 5 mEq/h, and adequacy of therapy must be checked further by endoscopy.

Zollinger-Ellison syndrome. Current concepts in diagnosis and management

The Zollinger-Ellison syndrome, although uncommon, is not rare, and most patients with the disorder present with clinical manifestations similar to those of patients with common peptic ulcer. Early studies emphasized death due to complications of massive gastric acid hypersecretion. However, with the availability of potent antisecretory agents to control acid secretion, death is now more frequently associated with the metastatic potential of slowly growing but malignant gastrinomas. Therefore, physicians should maintain a high degree of suspicion of the Zollinger-Ellison syndrome in assessing patients with either chronic peptic ulcer or unexplained secretory diarrhea. An evaluation aimed at early diagnosis of the Zollinger-Ellison syndrome should be instituted in such patients and should begin with a determination of the fasting serum gastrin level. At least 50 percent of patients with gastrinoma will have nondiagnostic serum gastrin concentrations and will therefore require provocative testing to establish the correct diagnosis. After the presence of the syndrome is established, patients should be treated with a potent antisecretory agent in doses sufficient to reduce basal acid output to less than 10 mmol in the hour preceding administration of the next dose. Although some patients may be maintained satisfactorily in this manner for extended periods, an approach aimed at tumor localization and extirpation is recommended in most patients. Preoperative evaluation should begin with CT scanning with intravenous contrast material. Selective angiography, and occasionally, portal venous sampling for gastrin, should be performed if the location and extent of tumor remain in question. If metastatic disease is demonstrated, or if MEN-I is present, surgery aimed at tumor resection, although it is occasionally effective, will probably be unsuccessful. Because of the considerable morbidity and mortality associated with pancreatoduodenectomy, it should not be performed for unresectable tumor in the head of the pancreas. In other patients with the Zollinger-Ellison syndrome, exploratory surgery should be performed; this should include a careful search for, and resection of, all pancreatic and extrapancreatic gastrinomas. With this approach, it is likely that at least 20 percent of all patients with the Zollinger-Ellison syndrome can be cured.

The use of high-resolution intraoperative ultrasound to localize gastrinomas: an initial report of a prospective study

We evaluated the use of high‐resolution, real‐time, B‐mode, intraoperative ultrasound (IOUS) compared to simple manual palpation to localize gastrinomas during laparotomies in 10 consecutive patients with Zollinger‐Ellison syndrome (ZES). All patients had clear biochemical evidence of ZES, and patients with metastatic gastrinoma identified preoperatively were excluded. Preoperative localization studies including ultrasound, computed tomography, and selective arteriography localized gastrinoma in 3 patients. Transhepatic portal venous sampling (PVS) for gastrin localized gastrin gradients to the pancreatic head region in 8 patients, and found no significant gastrin gradient in 2 patients. Gastrinomas were found at laparotomy and proven pathologically in 6 patients. Four patients had gastrinomas in the pancreatic head area as predicted by PVS; however, one patient had a tumor in the pancreatic tail and another in the jejunal wall. Palpation was the most sensitive intraoperative method to localize a gastrinoma (100% sensitivity). IOUS correctly imaged 3 gastrinomas (50% sensitivity). Palpation had 1 false‐positive finding (80% specificity) and IOUS had 2 false‐positive findings (60% specificity). All palpable masses that were also sonolucent on IOUS proved to be gastrinomas (3 patients). Thus, a combined positive finding raised the specificity and positive predictive value to 100%. We conclude that IOUS has utility in explorations for gastrinomas. Although not as sensitive or specific as palpation, IOUS can confirm, noninvasively, a suspicious palpable nodule. If IOUS fails to image a palpable nodule, our results suggest that the surgeon should still attempt to resect the nodule.

Localization and resection of gastrinomas in Zollinger-Ellison syndrome

From 1971-1986, 24 patients were diagnosed as having Zollinger-Ellison syndrome (ZES) and 22 patients had laparotomy. Of this group, gross tumor was identified in 15 of 22 patients. Ten of 15 patients had resection of their gastrinomas with the specific aim of curing the disease. This group had responded favorably to either cimetidine or ranitidine before operation. Preoperative transhepatic portal venous sampling (PVS) with gastrin determinations was performed in six patients; three patients had this procedure twice. The tumor was correctly localized by PVS in five of six patients. In four of six patients, the tumor was easily found at surgery. In two of six patients (33%) PVS was vital to intraoperative decisions. Criteria for biochemical cure are normal periodic fasting gastrin and secretin infusion tests. Of the 10 patients who had resection for potential cure, two patients failed within 48 hours of surgery on the basis of an elevated fasting serum gastrin level in one patient and a positive secretin infusion test in the other patient. Eight patients were considered cured with follow-up from 6 months through 15 years. Of the eight cured patients, the tumors were located as follows: four were extraintestinal and extrapancreatic, four were in the duodenal wall, one patient had a tumor located in the uncinate process of the pancreas, and one tumor was located in a lymph node along the lesser curve of the stomach. Two patients had mobilization of the pancreas and duodenum for a "blind" pancreatoduodenectomy based on preoperative PVS (2 procedures each patient). In one patient a 3-mm gastrinoma was enucleated from the posterior uncinate process. The second patient had pancreatoduodenectomy with findings of two duodenal wall gastrinomas. Both patients remained cured of ZES beyond 2 years. It is concluded that PVS does indeed locate some tumors before operation, even those not easily found at surgery. ZES can be cured by an aggressive approach combining preoperative tumor localization and tumor resection. Of the eight patients biochemically and perhaps biologically cured, follow-up was greater than four years in five patients, greater than two years in two patients, and beyond six months in one patient.