3.1 Secretory Studies 

It is easy to diagnose pancreatic insufficiency in the presence of the clinical triad of pancreatic calcification, diabetes and steatorrhea. Most pancreatic diseases, however, remain clinically silent until approximately 90% of the gland is destroyed. Lipase secretion appears to decrease earlier than trypsin secretion; hence, steatorrhea appears earlier than azotorrhea in patients suffering from pancreatic disease. Earlier recognition of pancreatic dysfunction may improve the management of the patient's disease and his or her quality of life.

Pancreatic function tests may be divided into two main groups: direct (duodenal intubation) and indirect (Table 1).

Tube tests require an oroduodenal tube that aspirates pancreatic secretion from the duodenum near the papilla of Vater so that the response to stimulating factors can be measured. The stimulants used are secretin, cholecystokinin and the Lundh test meal. The accuracy of these tests can be compromised by ineffective tube placement and lack of success in aspiration. This is partly compensated for by using measures of concentration of enzymes and bicarbonate. The collection period varies from 45 to 120 minutes.

The stimulation of the pancreas can be accomplished directly by infusing secretin alone or in combination with cholecystokinin. The combination allows the assessment not only of bicarbonate secretion (with secretin) but also of enzyme secretion, mainly trypsin.

A more physiological stimulation test of the pancreas by a meal is called the Lundh test. It assesses the response of the pancreas to endogenous secretin and pancreozymin (or CCK) released in response to a test meal of protein, fat and carbohydrates. The concentration of trypsin and the volume of secretion are measured in samples obtained in the duodenal aspirate. The Lundh meal is virtually always abnormal in pancreatic insufficiency. Unfortunately there is a borderline zone of abnormal values that are uninterpretable. In addition, many other factors influence the results of a Lundh meal, including small bowel mucosal disease, rate of gastric emptying, and surgical interruption of gastroduodenal anatomy. Although this is a more physiological test, its sensitivity and specificity are lower (70-80%) than those of direct hormonal stimulation.

Cannulation of the pancreatic duct during endoscopic retrograde cholangiopancreatography (ERCP) has been combined with direct stimulation of the pancreas. This technique allows the measurement of pure pancreatic juice uncontaminated by biliary or intestinal secretions, but this method is possibly no more sensitive than other tests in the diagnosis of pancreatic disease.

The intubation tests tend to be unpleasant for patients; they are also time-consuming and expensive and are performed mostly in specialized centers. Indirect tests of pancreatic function detect the result of pancreatic disease. The standard indirect test is the 72-hour fecal fat determination. The patient is placed on a 100 g/day fat diet and the stool is collected daily for three days. Individuals with normal pancreatic functions excrete less than 7% of the total amount of fat ingested, whereas those with pancreatic insufficiency excrete more than 20%. Although steatorrhea occurs in mucosal malabsorption, it is not as great as that encountered with pancreatic insufficiency. Measurements of stool nitrogen and stool chymotrypsin have not proved superior to fecal fat determinations. The major drawbacks to stool fat estimations are the lack of specificity and the inconvenience of collecting and analyzing the specimens. Attempts to screen for steatorrhea with less offensive tests (such as urine oxalate levels, ¹⁴C-triolein/³H-oleic acid assimilation test tripalmitate or palmitic acid breath tests) are promising but not generally accepted. After a rice-flour challenge, breath hydrogen is negligible in normal individuals but is dramatically increased in those with pancreatic insufficiency, in whom it is in turn reduced when the challenge rice flour is given with pancreatic enzymes.

There are two oral function tests available for assessing pancreatic functions: the bentiromide test and the pancreolauryl test.

The bentiromide test is useful in distinguishing patients with pancreatic steatorrhea from those with normal fat absorption. Bentiromide, a synthetic compound attached to para-aminobenzoic acid (PABA), is hydrolyzed by pancreatic chymotrypsin in the duodenum. This yields a low-molecular-weight substance, para-aminobenzoic acid, which is absorbed in the proximal small bowel and is partially conjugated in the liver. Metabolic byproducts of PABA are excreted in the urine. The excretion of less than 50% of the ingested dose in six hours indicates pancreatic exocrine insufficiency. Thus, the urine output of PABA should reflect duodenal chymotrypsin activity. Falsely abnormal results occur in patients with intestinal mucosal, liver or renal disease as a result of abnormalities of absorption, conjugation or excretion of PABA. A two-stage test has therefore been proposed in which PABA excretion following bentiromide is compared with the urine recovery of an equivalent dose of free PABA given on a subsequent occasion. PABA may also be measured in plasma instead of urine, and the plasma test may be more reliable in identifying patients with pancreatic insufficiency. The greatest use of this test may be in excluding pancreatic disease as a cause of diarrhea, steatorrhea, or weight loss.

The pancreolauryl test, using fluorescein dilaurate, has been extensively evaluated in Europe. It can detect only severe pancreatic insufficiency. This test is rarely used.

Chronic pancreatitis may give rise to an abnormal Schilling test, but rarely causes B₁₂ deficiency. Vitamin B₁₂ is released from food by gastric hydrochloric acid. This B₁₂ is bound to an R factor that is present in the saliva and the gastric juices. In the upper intestine, pancreatic enzymes release the R factor from B₁₂, which is then bound to intrinsic factor; the complex is subsequently absorbed in the terminal ileum. The Schilling test is relatively simple, but unfortunately it is not predictably abnormal except in instances of obvious pancreatic insufficiency.

Differentiating pancreatic carcinoma from chronic pancreatitis can at times be difficult; many tests have been described to aid diagnosis, but none are of proven value. Assay of carcinoembryonic antigen (CEA) in serum or from pure pancreatic juice obtained during ERCP has not proved to be a useful discriminator. The pancreatic oncofetal antigen has proved to be of uncertain significance. Serum galactosyl II transferase activity has recently been shown to be a reasonably specific indicator of pancreatic carcinoma in some patients. A sophisticated assay, it is unlikely to be suited to widespread use.

Trypsinogen, a proteolytic proenzyme, is exclusively produced in the pancreas. This enzyme can be detected by radioimmunoassay. It is elevated during an attack of pancreatitis and in renal failure, and is decreased in severe pancreatic insufficiency, cystic fibrosis and insulin-dependent diabetes without exocrine insufficiency. The levels of trypsinogen in cystic fibrosis decrease with age if the pancreas is involved. Low levels are found in about 60% of patients with pancreatic insufficiency. Patients with pancreatic insufficiency who have ongoing inflammation may have normal or raised levels. This fact, in addition to low levels in non-insulin-dependent diabetes, casts some doubt on the usefulness of this test in diagnosing pancreatic insufficiency. It may be useful in patients with steatorrhea that is due to nonpancreatic causes.

When faced with a patient with hyperamylasemia, it is necessary to exclude disease involving many organs other than just the pancreas (Table 2).

Amylase is produced and released from a variety of tissues, including the salivary glands, intestine and genitourinary tract. Normal serum contains three types of isoamylases as identified by isoelectric focusing. The pancreatic gland secretes one amylase at an isoelectric point of 7.0 that constitutes 33% of the total serum amylase. The parotid secretes several isoamylases with isoelectric points of about 6.4 and 6.0. Electrophoresis on polyacrylamide gel can separate five isoamylases on the basis of electrode mobility. Amylases originating in the fallopian tubes, tears, mucus and sweat have the same mobility as salivary amylase. All amylases have similar molecular weight and amino acid composition, but vary in terms of their glycosylation or deamination.

Amylase is filtered through the glomerular membrane and is reabsorbed in the proximal tubule. In healthy individuals, the amylase clearance parallels creatinine clearance. During acute pancreatitis, there is an increase in amylase clearance as opposed to creatinine clearance. Although this ratio was once thought to be specific to acute pancreatitis, other conditions that produce hyperamylasemia (such as diabetic ketoacidosis, burns, renal failure and perforated duodenal ulcer) may demonstrate a similar elevation. Occasionally, the serum amylase may be markedly increased in the absence of pancreatic or salivary diseases, whereas the urinary amylase is normal. In this instance, one must suspect either renal disease or macroamylasemia. In the latter condition normal serum amylase is bound by an IgA globulin, forming a complex that is too large to be filtered by the glomerulus. Affected individuals have an elevated serum amylase and a low to normal urinary excretion rate.

Frequently physicians are faced with a patient who has no overt salivary gland disease but has hyperamylasemia and no specific abdominal findings. As a rule, the level of amylase in pancreatitis usually is elevated to greater than 3 times the upper limit of normal and returns to normal within 2 to 10 days. If the amylase continues to be elevated in the absence of pancreatic complications, other causes (such as malignancy and macroamylasemia) should be investigated.

A rapid rise and fall in serum amylase in a patient with abdominal pain suggests the passage of a stone through the ampulla of Vater. When the serum amylase remains elevated for several days, the gallstone disease is usually complicated by pancreatitis.

Marked hyperamylasemia has been observed in patients with metastatic disease with ovarian cysts and tumors, and ruptured ectopic pregnancy. Isoamylase analysis reveals that the amylase has the same electrophoretic mobility as salivary-type isoenzyme. Macroamylase consists mostly of salivary amylase complexed with globulins, being therefore too large to be filtered at the glomerulus. Therefore these individuals have elevated serum amylase and low urinary amylase, with a low amylase-to-creatinine clearance ratio.

While the amylase levels in serum and urine are usually used as a measure of acute pancreatitis, measurements of lipase may be more specific and sensitive than total serum amylase. The assay of lipase is as accurate as the pancreatic isoamylase assay, and is likely to replace the amylase assay. Measuring both offers no advantage. Amylase and lipase measurements are readily available clinically, whereas radioimmunoassays are still being developed for other pancreatic enzymes (such as trypsin, chymotrypsin and elastase). Their role in the diagnosis of pancreatic disease needs to be established.