Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
A number of tests of exocrine pancreatic function have been devised over the years, but the search for a simple and yet sensitive and specific test has been elusive until recently. We will discuss the evidence that determination of the faecal concentration of pancreatic elastase 1 with an enzyme-linked immunosorbent assay method fulfils these criteria and thus provides an ideal tool to assess exocrine pancreatic function.
Approximately 90% of the pancreas is made up of exocrine tissue comprising acinar cells, which secrete pancreatic enzymes, and ductal cells, which secrete bicarbonate rich fluid. About 1% of the pancreas is endocrine tissue, the remainder being stromal tissue. The exocrine pancreas has a large functional capacity. More than 98% of function is usually lost before symptoms and signs of fat malabsorption are manifest.
The most common disease causing exocrine pancreatic insufficiency in children is cystic fibrosis (CF). Other conditions are rare but include chronic and familial pancreatitis, Shwachman–Diamond syndrome, Johanson–Blizzard syndrome, Pearson syndrome and isolated specific enzyme deficiencies—for example, lipase. A secondary pancreatic insufficiency may be seen in small bowel enteropathies, significant protein-energy malnutrition or short gut syndrome.1 There is increasing recognition of exocrine pancreatic dysfunction in children with diabetes.2
The clinical consequences of inadequate exocrine pancreatic output are mainly due to maldigestion of protein and fat rather than carbohydrate. With protein maldigestion, hypoproteinaemia and oedema may occur, and in fat maldigestion, overt malnutrition with steatorrhoea and wasting is more prominent. In practice, both tend to occur together and the signs and symptoms associated with the fat maldigestion are more obvious. Fat-soluble vitamin deficiencies also develop quickly.
Previously available pancreatic function tests
These tests measure the enzyme activities, the bicarbonate levels and the pancreatic juice flow rates in the duodenal juice after exogenous hormonal stimulation of the pancreas, usually with secretin and cholecystokinin or, occasionally, after a predetermined test meal (the Lundh meal). Such direct pancreatic function tests were considered the criterion standard for examining the exocrine function of the pancreas.3,–,5 However, there are major practical disadvantages. They require fluoroscopic duodenal intubation, are time consuming, are uncomfortable, are non-standardised and are expensive. There have been intermittent supply problems with secretin and cholecystokinin, and in addition, laboratory analysis of the enzymes has always been limited to a few academic units; hence, for routine clinical practice, direct pancreatic function tests are not a practical option.
Tubeless tests have included the fluorescein dilaurate (pancreolauryl) test and the N-benzoyl-tryrosyl para-aminobenzoic acid test, which rely on estimation of substrates in urine after ingestion and pancreatic digestion and therefore give indirect estimations of exocrine pancreatic function.6 7 Breath tests measuring radioactive or stable carbon isotopes after ingestion of labelled substrates have also been used. These tests are all somewhat cumbersome for use in children. Direct measurement of other pancreatic enzymes has been undertaken in faeces (chymotrypsin) and in serum (lipase) but lack specificity. Formal faecal fat collections for periods between 3 and 5 days with estimated fat intake for calculation of the coefficient of fat absorption are unpopular with children, parents and laboratories alike.
In summary, in addition to the practical issues with these tests in children, most of them have limited sensitivity in mild and moderate exocrine pancreatic insufficiency and may be interfered with by some drugs (including exogenous porcine pancreatic enzymes), diarrhoea, changes in intraluminal gut pH and gastrointestinal surgery.8
The ideal pancreatic function test, therefore, should be non-invasive, quantitative, reproducible, economic and easy to perform. Exogenous enzyme supplements should not interfere with the test performance.
Human elastase is synthesised by the acinar cells of the pancreas with the other proteolytic enzymes. It is composed of 240 amino acids. The molecular weight is 26 kd, and it has a special affinity for the carboxyl groups of alanine, valine and leucine.9 10 It was first described in 1950 and further characterised by Mallory and Travis as protease E in 1975.11 12 Largman et al13 in 1976 demonstrated its elastolytic properties.
The ontogeny of pancreatic elastase is similar to other exocrine proteolytic enzymes, and its decline in developing pancreatic insufficiency seems to mirror that of the other enzymes. Faecal elastase 1 (FE1) levels in meconium are low (<200 μg/g of stool) regardless of gestational age. Normal adult concentrations >200 μg/g of stool are reached by day 3 of life in term infants and by 2 weeks of age in premature infants of ≤28 weeks' gestation. After 2 weeks of life, whatever the gestational age, 96.8% of the infants with normal pancreatic function have FE1 levels comparable with those of adults.14
Under normal conditions, elastase 1 concentration in pancreatic juice is between 170 and 360 μg/ml, representing about 6% of all secreted pancreatic enzymes.15 During intestinal passage, elastase 1 is mainly bound to bile salts and not degraded. It becomes concentrated approximately fivefold to sixfold because water is reabsorbed, making it easier to measure in stool.15 16 It is stable at a wide range of pH and temperature—even storage for 5 days at room temperature does not influence immunological quantification in the stool—and so samples are easy to transport to laboratories capable of performing the analysis without special handling, provided that post office regulations for the transport of potentially infectious material are adhered to (rigid leak-proof container with sufficient absorbent wrapping to absorb leakage if it did occur and labelled “biohazard” or “pathology sample”). Unlike faecal chymotrypsin estimation, there is no cross reaction with porcine-derived elastase.
FE1 is measured by a sandwich enzyme-linked immunosorbent assay kit (ScheBo Biotech UK Limited, Basingstoke, UK) with two monoclonal antibodies that are highly specific for human pancreatic elastase 1, binding to two distinct epitopes. Results are reported as microgram per gram of stool. The lower detection limit is 15 μg/g.17 The cost of analysis is approximately £26, and turnaround time is approximately 21 days.
A single-spot stool sample of cherry size (approximately 100 mg) is sufficient. Watery stool samples are not recommended because FE1 levels may be falsely lowered by dilution. However, a semiliquid sample usually gives a reliable result. Samples are stable for convenient mailing and may be stored in the laboratory for up to 3 days at 4°C–8°C or for up to 1 year at −20°C.
Most of the evaluation in children comes from studies in CF, in other words, children with moderate or more commonly severe pancreatic insufficiency. However, the sensitivity in mild exocrine pancreatic insufficiency is much better than other indirect tests.10 In 204 children with CF, Terbrack et al18 found a sensitivity of 89.5% and a specificity of 99% for FE1 concentration, with a cutoff of <200 μg/g of stool.
Beharry et al observed an FE1 level of above 200 μg/g of stool in all their control group of 57 children with no steatorrhea determined by formal estimation of coefficient of fat absorption. All 39 patients with steatorrhea caused by CF had FE1 concentrations <100 μg/g of stool. Sensitivity and specificity in patients with CF were 100%. Another 10 patients with steatorrhoea from a variety of conditions (including the rare Shwachman–Diamond and Johanson–Blizzard syndromes, subtotal pancreatectomy and fibrosing pancreatitis) had FE1 levels below 100 μg/g. One patient with chronic pancreatitis had a level of 150 μg/g. The sensitivity of the test to detect primary insufficiency overall was 98%, using a cutoff of 100 μg/g of stool. The specificity of the test in determining the cause of steatorrhea (pancreatic maldigestion vs intestinal malabsorption) was 80%.19
In 1996, Loser et al compared FE1 determination with the direct secretin–caerulein test and faecal estimation of chymotrypsin levels in 79 patients with clinically suspected chronic pancreatitis and 50 healthy controls. Using an FE1 cutoff <200 μg/g of stool, he showed sensitivity of 93% and specificity of 93% for the detection of pancreatic insufficiency. At the same time, the sensitivity of faecal chymotrypsin was 64% with a cutoff <3 U/g of stool. In severe cases with steatorrhoea, the sensitivity was 100% and the specificity was 96%. In mild exocrine pancreatic insufficiency, the sensitivity overall was 63%. This study also showed very low individual day-to-day variations of faecal elastase concentrations.10
Interpretation of results
Adults and children after the first month of life
Values <100 μg elastase per gram of stool are consistent with exocrine pancreatic insufficiency. In established pancreatic insufficiency, values >100 μg/g of stool are highly unusual.
Values between 100 and 200 μg elastase per gram of stool are suggestive of exocrine pancreatic insufficiency and should be interpreted in the light of the clinical circumstance. It is reasonable to confirm the result with a second sample and pursue further with a formal faecal fat collection in the first instance. Such results may also be an indication of small bowel enteropathy (see below).
Summary of advantages of FE1 measurement:
▶. FE1 is pancreas specific.
▶. Because FE1 is stable during intestinal transit, the stool concentration accurately reflects the secretory capacity of the pancreas (diagnosis or exclusion of pancreatic exocrine insufficiency).
▶. FE1 determination correlates well with direct studies. It is more sensitive than faecal chymotrypsin.10
▶. Intraindividual variation of FE1 concentration is low.10
▶. The test is not affected by previous gastrointestinal surgery or gastric dysmotility.
▶. The monoclonal antibodies used in the test do not cross react with elastases of animal origin, so digestive enzyme substitution therapy has no influence on the determination of FE1 (unlike faecal chymotrypsin).
▶. The high stability of FE1 allows time for convenient mailing of samples.
There are a number of reasons to think about possible pancreatic pathology in clinical practice. The child who presents with persistent poor weight gain despite the demonstration of adequate nutritional intake may be malabsorbing because of a small bowel enteropathy but, equally, might have pancreatic insufficiency. The presence of abnormal stools increases the risk of identifiable pathologic conditions with or without an accompanying history of respiratory symptoms. The stool in exocrine pancreatic insufficiency tends to be more overtly greasy than that resulting from small bowel enteropathy. The consistency is soft rather than watery, and the colour is often pale. Parents may comment on the offensive smell.
Although it is true to say that in most children with CF, exocrine pancreatic function is abnormal by the time of diagnosis, in some circumstances, particularly in those diagnosed by neonatal screening, pancreatic function will decline over time. This group may be completely asymptomatic initially, but vigilance will allow appropriate nutritional intervention at an early stage before malnutrition occurs. Therefore, in children who are felt to be pancreatic sufficient at diagnosis, regular elastase monitoring is justified and should certainly be performed in those with faltering growth or abnormal stools.
Paediatricians see many children with episodic abdominal pain. Clearly, most do not have pancreatic pathologic conditions, but it can present with recurrent pain alone. Therefore, if the pattern of symptoms is in anyway unusual (including localised pain or periodic vomiting) or more severe than the norm, it is sensible to exclude a potential pancreatic pathologic condition.
Summary of indications:
▶. possible steatorrhoea
▶. faltering growth patterns
▶. chronic diarrhoea
▶. evaluation of pancreatic function in patients with CF
▶. diagnosis/exclusion of pancreatic involvement in recurrent abdominal pain—for example, possible chronic/recurrent pancreatitis
▶. follow-up of patients with known mild to moderate pancreatic insufficiency—possible demonstration of longitudinal decline in exocrine pancreatic function.
Small bowel mucosal integrity is necessary for normal control of exocrine pancreatic secretion. In small bowel enteropathies of any cause, faecal elastase levels may be low despite normal pancreatic functional capacity. This probably relates to alteration of enteric hormonal release, particularly cholecystokinin, and reverts to normal once the enteropathy has improved. There is also a transient fall in acute diarrhoeal illness. This highlights the need for consideration of mucosal biopsies in individuals with reduced FE1 concentrations in whom causes of primary pancreatic insufficiency have been excluded. FE1 concentration, therefore, also acts as a potential marker of small bowel enteropathy.1
The test is less sensitive in mild exocrine pancreatic insufficiency, and values between 100 and 200 μg/g of stool demand further investigation. Examples of relevant conditions include chronic or recurrent pancreatitis. Where the test result is borderline and/or clinical suspicion is high, further investigation including formal estimation of coefficient of fat absorption, estimation of serum levels of vitamins A, D and E and pancreatic imaging is warranted. Such results may also add impetus to any debate about the need for small bowel biopsy.
Low FE1 levels in children with Shwachman–Diamond syndrome should be interpreted with some caution because in this condition, it is possible to have low FE1 levels with adequate functional capacity for fat digestion; in other words, these children remain clinically pancreatic sufficient for longer because changes in specific enzyme activities occur at differing rates.20
Exclusion of meat from the diet has been demonstrated to result in significant changes in pancreatic secretion with a selective decrease in FE1 output. This may relate to a higher intake of soya trypsin inhibitors reducing proteolytic activity. Another possible mechanism is that high-fibre intake is known to affect pancreatic enzyme activity in vivo and in vitro, possibly because the colonic microflora is affected and bigger stools produce a dilution effect. Selenium deficiency in the vegetarian diet is also known to reduce exocrine pancreatic function.21 22
Is estimation of FE1 levels in children who present with failure to thrive indicated as a first-line investigation?
Pancreatic insufficiency is not a common cause for failure to thrive in children. Often, nutritional and gastrointestinal history and physical examination guide us to a reasonable clinical diagnosis.
In children who present with bulky and greasy stools and weight loss despite adequate nutritional intake or if there is chronic diarrhoea in the absence of serological evidence of coeliac disease and without any suggestion of food protein intolerance, one should consider a pancreatic pathologic condition. In these selected groups, it is appropriate to request estimation of FE1 levels.
If pancreatic insufficiency is then suggested, one should investigate for the cause. CF is by far the commonest reason for exocrine pancreatic failure in children.
In children who are diagnosed as having CF and who do not have any gastrointestinal symptoms, does routine estimation of FE1 levels identify pancreatic insufficiency early?
CF is a multiorgan disease. The clinical presentation varies. In some children, exocrine pancreatic failure is the predominant problem. Even if gastrointestinal symptoms are not prominent, these children may already have pancreatic insufficiency and, if not, are highly likely to develop it. The sooner it is diagnosed, the sooner appropriate management is instituted and the less likely these children are to become nutritionally compromised. Chronic respiratory illness, recurrent acute infections, recurrent hospital admissions and poor nutritional intake all contribute to malnutrition, and early clinical evidence of pancreatic failure can be missed, leading to a worsening of prognosis.
In these cases, it is very useful to judge the functional capacity of the pancreas even if there is no history of chronic diarrhoea or bulky and greasy stools. Properly adjusted enzyme replacement therapy often helps to restore the nutritional status of these children even before they become symptomatic.
Is treatment with enzyme replacement indicated in children with borderline FE1?
Borderline results for FE1 level should be interpreted with caution. The test should be repeated to make sure there is no sample error. If the repeat test result is also between 100 and 200 μg/g, it suggests that there is some element of exocrine pancreatic dysfunction.
If the child is symptomatic with failure to thrive and chronic diarrhoea, it is prudent to do further investigations to establish the cause for the abnormal result before considering enzyme therapy. This includes investigation for CF, and the other potential causes of exocrine pancreatic failure as indicated.
Any small bowel enteropathy including coeliac disease may give rise to a minor degree of exocrine pancreatic insufficiency caused by interference in the enteric hormonal axis due to small bowel mucosal damage. In this group of children, once the small bowel disease is treated, pancreatic function improves.
In a small group of these patients, when there is no obvious pathologic feature in the small bowel, the pancreas is structurally normal and findings of other investigations have been negative, it is reasonable to start (and monitor) enzyme replacement therapy to improve the nutrition of the child. This will also improve the symptoms of malabsorption.
In children with a proven pancreatic insufficiency syndrome—for example, Shwachman–Diamond syndrome, is testing of FE1 levels useful? Do these children need enzyme replacement?
In these circumstances, with a proven diagnosis, FE1 levels give very useful information in planning enzyme replacement therapy in conjunction with nutritional markers including fat-soluble vitamins.
However, we also know that these children can maintain pancreatic function despite low FE1 levels because of parallel development of other proteolytic enzymes, so treatment strategy should be based on the investigation results and the clinical symptoms and signs.
In children with recurrent abdominal pain, does a normal FE1 level rule out a pancreatic pathologic condition?
In the common pattern of non-specific recurrent abdominal pain in otherwise healthy young children, there is probably no reason to do anything else. Normal levels of FE1 mean that there is no exocrine pancreatic insufficiency.
However, in children with chronic pancreatitis who present with recurrent pain, the FE1 level may be normal if there is still adequate functional reserve. The diagnosis and the subsequent investigation here depends on clinical suspicion.
If FE1 levels are low in children who present with recurrent pain, it suggests a pancreatic pathologic condition. The possibility of chronic pancreatitis should be investigated with further imaging/endoscopic retrograde cholangiopancreatography.
One should also not forget the possibility of small bowel enteropathy, which may present with recurring abdominal pain.
Clinical bottom line: summary
▶. FE1 is a simple, non-invasive tool and is useful in the clinical assessment of children suspected of having pancreatic insufficiency, provided that the limitations of the test are recognised.
▶. A result >200 μg/g of stool reliably excludes exocrine pancreatic insufficiency.
▶. A result <100 μg/g of stool provides enough information to warrant a clinical trial of pancreatic enzyme therapy, especially if supported by other evidence of fat malabsorption—for example, estimation of serum levels of fat-soluble vitamins.
▶. In mild exocrine pancreatic insufficiency, faecal elastase is less sensitive but remains better than the other non-invasive tests.
One random semisolid to solid stool specimen weighing at least 100 g (cherry sized) should be sent to the laboratory in an ordinary container at room temperature. This specimen can be stored at 4–8°C in the fridge while waiting for transport.
▶. Growth faltering
▶. Chronic diarrhoea
▶. Pancreatic pathology.
Competing interests None.
Provenance and peer review Commissioned; externally peer reviewed.