Primary Hepatic Gastrinoma

The case of Zollinger-Ellison syndrome (ZES) reported by Evans and associates¹ contains a number of interesting points regarding ZES that are currently important in the management of this disease and that are seen frequently in some cases and infrequently in others. Two of these points will be discussed in this review: the long delay in diagnosis, which is not uncommon even today, and whether a primary hepatic gastrinoma, which is an uncommon occurrence, was likely present in this case.

The ZES patient in this case report had a 20-year delay in diagnosis, which is not uncommon even today, as the mean delay in diagnosis is currently 6 years.² Unfortunately, the time to diagnosis is not becoming shorter, even with the widespread availability of gastrin radioimmunoassay^2,3 and information from more than 3,500 articles written on ZES, including excellent reviews on diagnosis and other aspects of the disease. This delay continues to occur even though comprehensive analyses from long-term studies on ZES patients conducted by the National Institutes of Health, along with other analyses of the available world literature, have recently provided detailed guidelines for diagnosis. For the first time, guidelines derived from statistical analysis of large numbers of prospectively studied patients have become available. These guidelines include recommendations for when the disease should be clinically suspected²; when fasting gastrin levels should lead to suspicion of ZES⁴; when secretin and other provocative tests should be used; the formation of a definition of new criteria with enhanced sensitivity and specificity⁵; the description of acid secretory criteria for diagnosis⁶; and the reporting of specific features for recognizing ZES as part of the multiple endocrine neoplasia type 1 (MEN 1) syndrome (25% of cases).⁷ Unfortunately, the diagnosis of ZES is becoming more difficult, rather than easier, with the widespread use of proton pump inhibitors (PPIs).⁸ PPIs mask the clinical presentation of a patient by controlling ZES symptoms at conventional doses used for the treatment of patients with idiopathic gastroesophageal reflux disease/peptic ulcer disease (PUD) and also complicate the diagnosis by causing hypergastrinemia secondary to achlorhydria/ hypochlorhydria in patients without ZES.^8,9 It is, therefore, important to clinically suspect ZES in any patient presenting with PUD and diarrhea or associated endocrinopathies or any patient with familial PUD, severe and/or refractory PUD, or with prominent gastric folds found on endoscopy (which is present in 92% of ZES patients).^2,10 Although fasting serum gastrin (FSG) is the initial study usually performed in patients with suspected ZES, it is important to remember that PPIs are currently the most common cause of hypergastrinemia, that they often elevate FSG into the range that is commonly seen in ZES,^4,9,10 and that they can even result in false-positive secretin tests.¹¹ Therefore, it is essential to determine that the gastric pH is less than 2 at the time of hypergastrinemia to firmly establish the diagnosis of ZES, if the FSG is elevated greater than 10-fold,^3,4 as in the present case. If the FSG is elevated less than 10-fold, a secretin test and basal acid study should be performed.^3-5 No level of elevation of FSG on PPIs establishes the diagnosis of ZES.^3-5

A second important point raised in this case¹ is the difficulty in establishing the location of a primary gastrinoma in a non–duodenal-pancreatic site. In the literature, along with the usual location of the duodenum or pancreas, gastrinomas are reported to occur in other intra-abdominal locations (ovary, mesentery, lymph nodes, stomach, liver, biliary tract, gallbladder, renal capsule, pylorus, splenic hilium),^12,13as well as outside the abdomen (heart, nonsmall cell lung cancer).^14,15 Unfortunately, the location of a primary gastrinoma can be difficult to establish because gastrinomas frequently metastasize to the liver or lymph nodes in various adjacent sites and can invade them^16-19; therefore, the finding of a gastrinoma in a non–pancreatic-duodenal location can be either primary or a metastasis. Unfortunately, duodenal primaries, in particular, can be small and missed on all imaging studies as well as at surgery,^16,20 particularly if a duodenotomy was not performed.^16,20 This makes it difficult to exclude a small missed primary (particularly a duodenal primary) when only an extra–pancreatic-duodenal gastrinoma is found, as in the present case. The only method to clearly resolve this difficulty at present is to follow the patient post-resection and determine whether they are disease-free.²¹ Studies show that post-noncurative resection, FSG can return to normal and the detection of persistent disease requires both assessment of FSG and secretin test postoperatively at regular intervals.^17,21 If both FSG and the secretin test remain negative at 1 year, the patient has a 95% probability of being cured 3 years post-resection.²¹ At present, more than 85% of primary gastrinomas are found in the duodenum (60–80%) or pancreas (10–40%).^10,12,16 Although lymph node primaries are controversial, they have been reported to occur in up to 10% of cases¹⁷ in the National Institutes of Health studies. Ovarian primaries have also been reported to occur in 0.5% of cases in these studies. Fewer than 30 primary hepatic or biliary tract gastrinomas have been reported, and we have found that they have occurred in less than 2% of National Institutes of Health ZES patients.^{3,12,16,18,19} As Evans and associates pointed out in their case discussion,¹ although a primary gastrinoma appears to be a strong possibility in their patient, the diagnosis cannot be definitively established for a number of reasons. First, a duodenotomy, which is required to identify small duodenal gastrinomas,^16,20,22 was not performed¹; thus, this possible primary location cannot be excluded. Second, both a FSG and a secretin test should be performed at each follow-up to detect all persistent/recurrent disease,²¹ as well as imaging studies at least at yearly intervals. At the first follow-up in this patient, only a FSG was performed, which misses 45% of patients with persistent/recurrent disease post-resection.²¹ Third, the follow-up in this case (2 months)¹ is much too short to establish a possible cure, as elevations of FSG/ secretin test are frequently not seen until more than 6 months after large noncurative resections.^17,21

The attempt to clearly establish the primary site in this patient is not merely an idle exercise, as this finding possibly has important prognostic information for the patient and physician. Our prospective studies have shown that any metastatic involvement of the liver, even if limited to 1 hepatic lobe, is associated with decreased survival.²³ In contrast, if the patient has a hepatic primary tumor and remains disease-free, their long-term survival approaches that of a normal patient. Furthermore, the need for additional antitumor treatment, the frequency of follow-up, and the extensiveness of tests required during follow-up are all influenced by the category in which the patient ultimately ends up.

Lastly, this case¹ demonstrates the importance of careful imaging in patients with ZES and the importance of considering surgical resection in all patients without MEN 1, in which the tumor may be resectable, even if it is involved to a limited extent in the liver.^3,16,19 Tumor progression is presently the single main cause of death in patients with ZES, now that gastric acid hypersecretion can be controlled medically in almost every patient.²³ Only surgery can possibly cure the patient and should be considered in all patients without MEN 1, patients without a concurrent illness that limits life expectancy, or in patients without nonresectable metastatic disease.¹⁶ In 5–15% of patients with liver metastases, the tumor is limited to a single liver lobe¹⁸ and in a small percentage, a liver primary may be present. Thus, in this group of patients, after careful imaging, if the disease is limited in its extent to the liver and is potentially completely resectable, resection should be considered.¹⁸ This is particularly true because recent studies show that surgical resection can decrease the possibility of developing liver metastases and that it increases survival in ZES patients.^24,25