In 1955, Zollinger and Ellison1 described two patients with jejunal ulceration, gastric acid hypersecretion, and a non-β pancreatic islet cell tumor. The diagnostic triad of gastric acid hypersecretion, recurrent peptic ulceration in the setting of adequate therapy, and a non-β islet cell tumor of the pancreas was proposed for the syndrome that now bears their names. The subsequent development of a sensitive radioimmunoassay resulted in the identification of gastrin overproduction as the underlying mechanism for this syndrome.

The incidence of gastrinoma is between one and three per 1 million people.2,3 In patients with peptic ulcer disease (PUD), gastrinoma is thought to account for 0.1% to 1.0% of cases. However, these figures may be underestimates because these patients may have symptoms similar to those with typical PUD, attributed to Helicobacter pylori or the use of nonsteroidal antiinflammatory drugs (NSAIDs), and may go undiagnosed. Enteropancreatic neuroendocrine tumors are second in incidence only to insulinoma.

Although 75% to 80% of gastrinomas are sporadic, 20% to 25% are diagnosed in the setting of multiple endocrine neoplasia type 1 (MEN1).2,3 This syndrome includes primary hyperparathyroidism and tumors of the pituitary gland in addition to pancreatic islets cell tumors and duodenal gastinomas. (Thymic carcinoids, adrenal adenomas, and some cutaneous tumors have also been associated.) This distinction is important to make because the diagnostic and management approaches to sporadic and familial disease differ significantly. The penetrance of pancreatic neuroendocrine tumors or duodenal gastrinomas in patients with MEN1 is approximately 60% to 70%. Gastrinoma is the most common of these tumors. In general, these tumors become clinically evident in the fourth or fifth decade of life in these patients.

These tumors arise from enteroendocrine cells found in the pancreas and duodenum. Histologically, they are well-differentiated tumors that resemble other pancreatic neuroendocrine tumors and demonstrate few mitoses with a proliferative (Ki-67) index of 2% to 10%. As with all gastroenteropancreatic neuroendocrine tumors, gastrinomas are classified by World Health Organization guidelines into well-differentiated endocrine tumors (benign), well-differentiated endocrine carcinomas (low malignant potential), and poorly differentiated endocrine carcinomas (high malignant potential).4 The majority of gastrinomas (50% to 80%) fit into the well-differentiated endocrine carcinoma category. Ultimately, more than 60% of these tumors prove to be malignant. This distinction relies on evidence of local invasion or metastatic spread because no histologic feature exists that is diagnostic of malignancy.

Although initially described as a pancreatic neuroendocrine tumor, it is now recognized that the majority of gastrinomas (50% to 88%) arise in the duodenum.2,3 In addition, these tumors are rarely found in the stomach, jejunum, biliary tract, liver, kidney, and ovary. Sporadic tumors of the duodenum are typically small (<1 cm). Despite this, 60% to 80% have metastasized to regional lymph nodes at the time of diagnosis. On occasion, gastrinoma is found in a lymph node when no primary tumor is evident. Most, if not all, of these “primary lymph node gastrinomas” have likely spread from a small duodenal tumor. With careful intraoperative examination, most of these primary tumors can be found. The existence of true lymph node primary tumors is currently controversial. The most common site of metastatic spread is the liver; the second most common is bone. In sporadic duodenal gastrinomas, liver metastases are seen in 10% of patients at the time of diagnosis.

In contrast to tumors in the duodenum, sporadic pancreatic gastrinomas are generally larger (>2 cm). Although spread to regional lymph nodes appears to occur at a similar rate to duodenal tumors, pancreatic gastrinomas are associated with a higher likelihood of distant liver metastases. Tumors associated with MEN1 are most commonly found in the duodenum (90%). Multiple tumors are the rule in these patients. Most of these tumors are small, and distant metastases are rare at the time of diagnosis.

Few specific risk factors have been identified for gastrinoma. Perhaps the only widely acknowledged risk factor is family history in those with MEN1. More recently, it has been proposed that prolonged hypergastrinemia from antisecretory medications or other causes may increase the risk of developing not only gastric carcinoids but duodenal gastrinomas as well. This is postulated to be a result of long-standing G-cell stimulation.

The clinical presentation of patients with gastrinoma is relatively nonspecific. Roy et al.5 from the National Institutes of Health (NIH) conducted a prospective evaluation of 261 patients with gastrinoma seen over a 25-year period. The mean age of presentation in this cohort was 41.1 ± 0.7 years. On average, patients with sporadic disease presented later than those with MEN1-associated gastrinomas. The most common complaint in this cohort was abdominal pain, which was seen in 75% of patients. Diarrhea was noted in 73% of patients, with symptoms of heartburn (44%), nausea (33%), vomiting (25%), and weight loss (17%) occurring in a lower percentage of patients. Abdominal pain was noted less frequently in those with MEN1-associated disease (66%).

Biochemical Tests

Given its nonspecific presentation and relative rarity, it is not surprising that the diagnosis of gastrinoma is often significantly delayed. In fact, the mean delay from symptom onset to diagnosis is 5.2 years. This may be even longer in patients taking proton pump inhibitors (PPIs). Clinicians should have a heightened index of suspicion in patients with recurrent or severe cases of PUD, multiple ulcers, or ulcers in unusual locations (more distal small bowel). In addition, in patients with duodenal ulcers in the absence of H. pylori infection, PUD associated with prolonged diarrhea, and PUD in the setting of personal or family history of hypercalcemia, the possible diagnosis of gastrinoma should be considered.

In the appropriate clinical setting, the diagnosis of gastrinoma is made with biochemical testing. The biochemical hallmark of gastrinoma is an elevation of fasting serum gastrin (FSG) level in the presence of gastric acid hypersecretion. The demonstration of a low gastric pH is important given that a number of conditions with associated achlorhydria (or hypochlorhydria) manifest an elevated serum gastrin level (Table 18-1). These conditions include pernicious anemia, atrophic gastritis, PPI or H2-blocker use, and postvagotomy states. Even in the setting of elevated gastric acid secretion, hypergastrinemia may result from other conditions, such as gastric outlet obstruction, antral G-cell hyperplasia, renal failure, and short bowel syndrome. Patients taking antisecretory medications should discontinue them at least 1 week before the FSG level is measured. An algorithm for the diagnosis of gastrinomas is shown in Figure 18-1. Given the broad differential diagnosis and the confusion regarding the interpretation of diagnostic testing, patients referred to an endocrine surgeon with a provisional diagnosis should be reviewed thoroughly because many of these diagnoses may be incorrect.

The interpretation of FSG levels has been the focus of some discussion in the literature.3 Overall, it is generally accepted that an FSG level more than 10 times the upper limit of normal in the setting of a gastric pH below 2 is diagnostic of gastrinoma. In our laboratory, the upper limit of normal is 47 pg/mL. Unfortunately, in a significant portion of these patients, the elevation in FSG level is lower than this benchmark. In an effort to clarify this, a recent NIH study examined prospective data on 309 patients with Zollinger-Ellison syndrome and compared these results with 2209 patients in 513 series reported on in the literature.6 In this study, a normal FSG level was seen in just 3% of patients with gastrinoma. If repeat testing was used, this number decreased to 0.3%. In addition, two thirds of these patients were found to have levels below 10 times the upper limit of normal.

Patients with suspected gastrinomas who have elevated FSG levels less than 10 times the upper limit of normal should undergo provocative testing. Secretin and calcium stimulation tests have been described for this disease. Although the mechanism remains unclear, secretin stimulates gastrin release in patients with gastrinoma but inhibits its release in those without this tumor. This test involves the administration of a 2-U/kg intravenous bolus infusion of secretin with FSG measurements at specific time points (−15, −5, 0, +2, +5, +10, +15, +20 and +30 minutes) before and after infusion. Similarly, calcium infusion has been noted to produce an increase in gastrin level in these patients. This test is performed by administering an infusion of 5 mg/kg/hr of calcium (typically in the form of a 10% calcium gluconate solution) over 3 hours with FSG measurements at 30, 15, and 0 minutes before and 30-minute intervals after the start of this infusion for 3 hours.