The role of surgery in the Zollinger-Ellison syndrome

Current management of Zollinger-Ellison syndrome

Patients with Zollinger-Ellison syndrome require that management decisions be made to control the gastric acid hypersecretion and treatment directed at the gastrinoma itself. The advent of newer antisecretory drugs and increased knowledge of the natural history of this disease have led to major changes in the management of each of these two areas. Recent studies have demonstrated that treatment with the currently available histamine H2-receptor antagonists (cimetidine, ranitidine) with or without an anticholinergic agent will control gastric acid secretion in almost all patients. These studies have also shown that most patients require higher doses than those used routinely to treat peptic ulcer, treatment is only successful if an adequate dose of antisecretory drug is used and must be monitored by measuring gastric acid hypersecretion, and established criteria to regulate the dose must be used. Newer more potent antisecretory drugs such as famotidine or omeprazole will facilitate management of gastric hypersecretion but are not yet currently available. Highly selective vagotomy should be considered in those patients who require high doses of cimetidine or ranitidine. Total gastrectomy should be reserved for those patients unwilling or unable to take oral medication. Although aggressive surgery is not warranted in most patients because overall prognosis is excellent, tumour status should be assessed in all patients by imaging studies (CT scan, ultrasound, selective angiogram). Patients without metastatic disease and without the MEN-1 syndrome (multiple-endocrine-neoplasia type 1) should undergo exploratory laparotomy by a surgeon experienced in treating this disease, with studies suggesting a cure rate of approximately 20%.(ABSTRACT TRUNCATED AT 250 WORDS)

Gastrinoma

Although there are multiple etiologies of clinical hypergastrinemia, the most important cause is gastrinoma [Zollinger-Ellison (ZE) syndrome]. The most significant advance in the management of patients with the ZE syndrome has been the development of the ability to establish firmly the correct preoperative diagnosis by radioimmunoassay. We believe that all patients who have the ZE syndrome should undergo operation for two reasons: to identify and resect all tumor and because in patients in whom no tumor is found or in whom all tumor cannot be resected, the most reliable relief of hypersecretory symptoms for the life of the patient can be achieved by total gastrectomy.

Ectopic gastrinoma and Zollinger-Ellison syndrome

In Zollinger-Ellison syndrome (ZES), the discovery of gastrinomas in unusual locations, such as a lymph node, poses a diagnostic problem centered on whether the neoplasm is primary or metastatic. The clinical, gross, microscopic, immunocytochemical, and ultrastructural features of ectopic gastrinomas were studied in four patients with ZES, and reports of 14 similar cases were reviewed. These extragastroenteropancreatic (EGEP) gastrinomas have many of the morphologic features of gastrinomas in conventional locations. However, the centrifugal expansile growth pattern, characterized by a thick fibrous capsule, hyalinized fibrous septa, and, frequently, cystic degenerative changes in EGEP gastrinomas should alert the pathologist to the probability that these neoplasms are primary. Additional evidence for the primary nature of these EGEP gastrinomas is derived from the postoperative normalization of high serum gastrin levels and the correction of the abnormal gastrin response to secretin challenge or to calcium infusion tests. Increased awareness of the occurrence and features of these EGEP gastrinomas is crucial for both pathologists and surgeons to ensure proper evaluation and treatment of patients with ZES.

Surgery for peptic ulceration associated with hypergastrinaemia

Between 1971 and 1983, 31 males and 13 females were found to have peptic ulceration associated with hypergastrinaemia. An antral G-cell lesion was present in 11 (25 per cent) and a gastrinoma in 14 (32 per cent). There were 11 patients with multiple endocrine adenomatosis (MEA) (25 per cent) and 4 (9 per cent) with primary hyperparathyroidism. Four patients (9 per cent) were unclassified. Length of history and level of gastrin did not differentiate between the groups and an average of 2.5 operations was performed per patient, while the overall mortality was 27.3 per cent. The patients with G-cell lesions were significantly younger than all the other groups (P less than 0.01). Partial gastrectomy adequately treated G-cell hyperplasia. Total gastrectomy was required to treat pancreatic gastrinomata but additional pancreatic resection did not improve the outcome. In MEA, parathyroidectomy did not influence the treatment of a gastrinoma. This is the first recorded experience of surgery for hypergastrinaemia in the United Kingdom and the outcome of such a retrospective study may be a guide to the future management of these conditions.

Present experiences with omeprazole in the Zollinger-Ellison syndrome

As effective tumour resection can be achieved in only a minority of patients with Zollinger-Ellison syndrome, therapy in such patients is usually directed to treatment and prevention of symptoms induced by hypersecretion of gastric acid. Omeprazole is a potent and long-acting inhibitor of pentagastrin stimulated gastric acid secretion in normal subjects, and this prompted a study of the effect of omeprazole on gastric acid hypersecretion and acid peptic disease resulting from endogenous hypergastrinaemia in Zollinger-Ellison patients. Omeprazole was found to be effective even in those patients whose symptoms were not adequately controlled by high doses of histamine H2-receptor antagonists, with or without pirenzepine. Serum gastrin levels were not affected; no laboratory abnormalities and minimal side-effects induced by omeprazole have been reported in Zollinger-Ellison patients even on long-term studies of up to 27 months.

Treatment of Zollinger-Ellison syndrome with exploratory laparotomy, proximal gastric vagotomy, and H2-receptor antagonists. A prospective study

Twenty-two patients with Zollinger-Ellison syndrome were managed by a combined medical and surgical approach. Patients were treated initially with cimetidine or ranitidine. A laparotomy was performed to remove easily resectable tumors and to carry out a proximal gastric vagotomy. Tumors were found in 9 patients (41%) and all visible tumors were removed from 6 of the 9 patients. Fasting serum gastrin concentrations and serum gastrin responses to intravenous secretin were normal 6 wk after surgery in each of the patients from whom all visible tumors were resected and are normal in 4 patients, 6 wk to 5 yr after surgery. Acid secretion was reduced after vagotomy in each patient, even when tumors were not found or completely resected. Thus, vagotomy decreased the acid secretory response to endogenous hypergastrinemia. In addition, vagotomy augmented the inhibitory effect of H2-receptor antagonists on acid secretion. Follow-up has ranged from 6 wk to 6 yr (median, 2 yr). Dosages of cimetidine or ranitidine have been reduced, compared with preoperative amounts, in all but 1 patient. Two patients are taking no antisecretory drugs. Only 3 patients have had occasional symptoms of ulcer disease. Complications such as bleeding, perforation, or obstruction have not occurred in any patient. Endoscopy was performed in all patients to estimate the point prevalence of active ulcers and an ulcer was found in 1 patient. Based on these results, it is our opinion that this combined medical and surgical approach is an effective treatment for patients with Zollinger-Ellison syndrome.

Benign and malignant gastrinoma

The advent of the histamine H2-receptor antagonists and the renewed interest in curative surgery in patients with gastrinoma have made the differentiation between benign and malignant tumors of critical importance. An analysis of 65 patients with gastrinoma followed for an average of 93 months revealed two distinct clinical groups: those with and those without hepatic tumors at initial examination or operation. Among the 14 patients with hepatic tumors, 12 had multiple liver metastases from pancreatic or duodenal primary tumors, and 2 had primary hepatic gastrinomas. Ten of the 14 patients (71 percent) died from tumor progression, and the total tumor-related mortality for this group was 79 percent. In contrast, only 1 of 15 patients (7 percent) with tumor in the lymph nodes died from a tumor-related cause (recurrent ulcer hemorrhage), and none died from tumor progression. Only a single patient with lymph node metastases at initial exploration went on to the development of liver metastases, which was found incidentally at autopsy 313 months later. Among 23 patients with either primary tumors only or no tumors found at laparotomy, there was only one tumor-related death and no deaths from tumor spread. Life-table analysis demonstrated a significantly decreased length of survival for patients with liver tumor compared with those without liver involvement. Multiple endocrine adenopathy syndrome was not a significant factor in survival. Serum gastrin levels were likewise nondiscriminatory. Six of 52 patients (12 percent), including three with tumor in the lymph nodes, were apparently cured by excision of all gastrinoma recognized at laparotomy. The cure rate was 23 percent for patients without multiple endocrine adenopathy syndrome or liver metastases. Hepatic metastases is a definitive marker for clinically malignant disease and portends a poor prognosis. Patients with gastrinoma confined to the lymph nodes uncommonly follow a malignant clinical course. Such patients have at least a 20 percent probability of surgical cure if they do not have multiple endocrine adenopathy syndrome.

The gastrinoma triangle: operative implications

Operative experience with 45 gastrinoma patients has led to the identification of an anatomic area where occult tumors can be discovered and where excision of these tumors had led to apparent cure. Of 36 patients with histologically confirmed gastrinomas, 27 patients (75 percent) had obvious and 9 patients (25 percent) had occult tumors. All nine occult lesions were found within an anatomic triangle defined by the junction of the cystic and common bile ducts superiorly, the junction of the second and third portions of the duodenum inferiorly, and the junction of the neck and body of the pancreas medially. Although occult tumors from three patients (in the lymph nodes in two patients and in the duodenum in one patient) were removed primarily for histologic diagnosis, postoperative serum gastrin levels have remained within the normal range (follow-up of 86 to 99 months). Two patients had excision of the tumor with intent to cure. One patient with a solitary duodenal tumor was apparently cured but committed suicide 3 months postoperatively. The other patient had both obvious primary and occult metastatic tumors within the triangle and was eugastrinemic 9 months after excision. In all patients in whom tumor was found, it was locally excised, and no patient was subjected to radical pancreatic resection. There were no postoperative complications related to tumor removal. An aggressive approach towards curative tumor excision is now advocated for all gastrinoma patients who are suitable operative risks and have no evidence preoperatively of liver metastases or the multiple endocrine neoplasm-type I syndrome.