Original Article Ann Clin Biochem 1999; 36: 739-742

Faecal elastase 1: a marker of exocrine pancreatic insufficiency

in cystic fibrosis
I J Phillips', D J Rowel, P Dewar? and G J Connett-
From the 1Department of Chemical Pathology (NHS) and the 2 Department of Paediatrics,
Southampton General Hospital, Southampton S016 6YD, UK

SUMMARY. Pancreatic elastase 1 (E1), a digestive protease, is synthesized by the

acinar cells of the pancreas. Using an enzyme-linked immunosorbent assay, we
evaluated stool El levels in the following groups of patients:
(a) Specimens submitted for occult blood examination from 20 adults, over 3
consecutive days, to assess the inter-day variability in E1 excretion. There were
no symptoms suggestive of pancreatic insufficiency in this group. The mean E1
concentration over all samples was 4571J.g E1/g stool (range 124-1683). The
intra-assay variation was 6-4% (n= 14) and the inter-assay variation was 8·8%
(n = 12). The mean intra-patient variation was 17%.
(b) Cystic fibrosis (CF) patients. Eight patients had E1 levels in the reference range
(> 200 1J.g E1/g stool). The remaining 25 patients had undetectable E1 levels.
(c) A control group of children presenting with unexplained bronchiectasis and/or
recurrent respiratory infections and no symptoms of pancreatic dysfunction. The
mean El concentration in the group was 519 1J.g E1/g stool (range 139-1941).
There was no significant difference in E1 concentrations between the two non-CF
groups, nor between the pancreatic-sufficient CF patients when compared with both
non-CF groups. There was a significant difference between the pancreatic-sufficient
and -insufficient CF groups (P < 0·001) using the Mann-Whitney U test.
All fifteen CF patients who were ~F508 homozygotes had undetectable E1. It may
be possible to relate CF genotype to the presence or absence of E1 and to the degree
of pancreatic insufficiency. Measurement of faecal E1 in children with CF appears to
differentiate them into a group of children with normal pancreatic function and a
larger group with severe insufficiency.

Additional key phrases: proteases; pancreas

Cystic fibrosis (CF) is a common, life-shortening,
autosomal recessive disorder caused by mutations
in the cystic fibrosis transmembrane conductance
regulator (CFTR) on chromosome 7. 1
Approximately 85% of those affected have
exocrine pancreatic insufficiency caused by
obstruction of the small ducts by viscous
secretions leading to necrosis of the acinar and
ductal cells and eventual fibrosis of the pancrea-
tic lobules. The loss of exocrine pancreatic
function leads to maldigestion and malabsorp-
tion. Evaluation of exocrine pancreatic function
Correspondence: Mr Ian Phillips.
E-mail: lan(a)PhiIl2046.freeserve.co.uk
is difficult. Current methods involve invasive
'direct testing' (e.g. of cholecystokinin-pancreo-
zymin, secretin-pancreozymin, and the Lundh
test) or 'indirect testing' (e.g. of p-aminobenzoic
acid, faecal chymotrypsin, and the pancreolauryl
test). Indirect tests of exocrine pancreatic
function are preferred for routine clinical use,
but have poor sensitivity in mild pancreatic
insufficiency, intestinal disease with malabsorp-
tion and liver disease with cholestasis.?
Human pancreatic elastase 1 (EI) was first
isolated by Mallory and Travis 3 in 1975 and was
originally termed protease E. A year later,
Largman et al. 4 isolated elastases 1 and 2. E1
was identical to protease E. Pancreatic EI is a

739
740 Phillips et at.

digestive protease synthesized by the acinar cells

of the pancreas and secreted into the duodenum 10000
via the pancreatic duct. El is synthesized as a
zymogen and the mature enzyme has a mol-
ecular weight of 28 kDa.
Faecal enzyme measurements are convenient
and easy to perform. The concentration of El in
faeces was found to be five times greater than
that in pancreatic juice, indicating that El is not
degraded during intestinal transit. A highly
sensitive enzyme-linked immunosorbent assay
(ELISA) for human duodenal and faecal El,
using two specific monoclonal antibodies, is now 1000
commercially available.
t::.
~
••
The aim of our pilot study was to assess the a

~
D
imprecision of the faecal El assay, to evaluate
the inter-day variability of El excretion and
measure El in children with and without CF to
"0
•
evaluate its use as an indicator of pancreatic 0
"lii
t::. •
insufficiency, and to relate El excretion to CF .e' t::.
t::.
•
genotype. ~

w
OJ
.3
MATERIALS AND METHODS Q)
III
100
III
"lii
Patients III
Cii
Ethical permission was obtained before com- n;
mencing the study. A total of 67 patients were oQ)
III
categorized into three groups. Group A con- LL
sisted of 20 patients (mean age 57 years, range 5-
80 years, 8 men, 12 women). These patients were
being investigated for the presence of faecal
occult blood (FOB). We used specimens pro-
vided over 3 consecutive days to evaluate the
10
w:»:1
'»»1
•
inter-day variation in El excretion.
~
Group B consisted of 14 children (mean age
8·25 years, range 2-19 years, 7 boys, 7 girls) with w:»:1
unexplained bronchiectasis and/or recurrent
respiratory infections and no clinical signs of W»7
any gastrointestinal pathology.
Group C consisted of 33 children with proven
CF (mean age 3·8 years, range 3 months to 7
years, 23 boys, 10 girls). Of the 33 children in
this group, 26 were judged to be pancreatic-
insufficient on clinical grounds including failure
2 3 4
to thrive and/or presence of steatorrhoea. The
remaining seven children were judged to be Group
pancreatic-sufficient. FIGURE 1. Faecal elastase 1 (£1) concentrations in
patients with cystic fibrosis (CF) compared with control
Methods groups. Group 1 = £1 concentration in samples sub-
All stool specimens were frozen at - 20°C prior mittedfor faecal occult blood analysis over 3 consecutive
to analysis for El. El was measured using an days (each symbol represents the mean of three
samples); group 2 = £1 concentrations in the control
ELISA (ScheBo'® Tech, Giessen, Germany)
group (children without CF); group 3 = £1 concentra-
employing two monoclonal antibodies binding tions in the pancreatic-insufficient CF group; group
to two distinct epitopes of this enzyme. One 4 = £1 concentrations in the pancreatic-sufficient CF
hundred millgrams of stool was diluted 1:500 in group.

Ann Clin Biochem 1999: 36

 Faecal elastase 1 741

sample wash buffer and the faecal elastase
concentration (/lgjg wet weight of stool) was
measured photometrically. The range of the
assay using the standards provided is 15-500 /lg
Eljg stool.
The manufacturers provide the following
clinical ranges:
Severe exocrine pancreatic insufficiency:
< 100/lg Eljg stool
Moderate exocrine pancreatic insufficiency:
1OG-200/lg Eljg stool
Normal levels: > 200 /lg E Ijg stool
RESULTS
The intra-assay imprecision of the stool EI assay
(14 duplicates of the same stool) was 6-4% at a
concentration of 282/lg E Ijg stool. The inter-
assay imprecision (12 duplicates of the same
stool specimen in 12 separate assays) was 8·8%
at a concentration of 205 /lg Eljg stool.
Figure 1 shows faecal elastase concentrations
in all of the patient groups. The results shown
for group A are the mean EI concentrations
from each set of triplicates. In group A the mean
E I concentration calculated from log trans-
formed data was 457/lg Eljg stool (range 124-
1683/lgjg). The mean within-individual varia-
tion was 18% for 20 triplicates. Two patients
had E I levels in the moderate insufficiency
range.
The mean faecal elastase concentration in
group B was 519/lg Eljg stool (range 139-
1941 /lgjg).
In group C, there were 26 patients who were
judged clinically to be pancreatic-insufficient. Of
these patients, 23 had undetectable EI levels
( < 15/lg EI jg stool), confirming severe pancrea-
tic insufficiency. The remaining three patients
had E 1 levels in the normal range ( > 200 ug E 1j g
stool), suggesting pancreatic sufficiency.
Of the seven CF children judged clinically to
be pancreatic sufficient, five had EI levels in the
normal range. However, two patients had
undetectable levels of elastase, suggesting pan-
creatic insufficiency. There was no significant
difference in elastase levels between the pan-
creatic-sufficient CF patients, the faecal occult
blood group and the non-CF children (P> 0,05),
using the Mann-Whitney U test. However, there
was a significant difference between the pan-
creatic-sufficient CF patients and the pancreatic-
insufficient patients (P < 0,001).
Table 1 shows elastase concentrations in
relation to CF patient genotype. All patients
homozygous for L\F508 (n = 15) had elastase
levels less than 15 /lg Eljg stool and were judged
clinically to be pancreatic-insufficient. CF pa-
tients who were heterozygous for the L\F508
mutation displayed varying elastase levels.
DISCUSSION
The objective of this study was to evaluate a
commercial ELISA assay for the measurement
of pancreatic EI in stool specimens from
children with and without CF and in stool
specimens received for faecal occult blood
analysis. The assay is highly specific for human
pancreatic EI. 5 Therefore, subjects do not have
to cease taking pancreatic supplements before
assessment of exocrine pancreatic function."
Pancreatic EI is highly stable in the intestinal
tract; its concentration in faeces is greater than
that in pancreatic juice and this reflects the
secretory capacity of the pancreas. The stability
of E 1 allows stool to be posted to reference
centres for analysis."
The El assay showed acceptable levels of
intra- and inter-assay imprecision. Analysis of
triplicate stool samples indicated acceptable day-
to-day intra-individual variation in EI excretion.

TABLE I. Faecal elastase I (E1) concentrations in children with cystic fibrosis. related to genotype

Elastase I concentrations (Jlg EI/g stool)

Number
Genotype of patients < 15 15-100 100-200 >200

i\F508/i\F508 15 15
i\F508/G551D I I
i\F508/G542X 3 3
i\F508/RI17H 3 3
i\F508/DiIOH I I
V520FjV520F I I
i\F508/Unknown 7 5 2
None identified 2 I I

Ann Clin Biochem 1999: 36

742 Phillips et at.
Two patients from group A had elastase I
levels in the moderate insufficiency range; the
remainder, including all the non-CF children,
were within the reference range. Eighty-eight per
cent of the CF children in the group judged
clinically pancreatic-insufficient had undetect-
able levels of EI. Of these, 45% were homo-
zygous for the AF508 mutation. This mutation is
the commonest mutation in Western Europe, in
which a three-base pair sequence coding for
phenylalanine is deleted at position 508 of the
CFTR protein. Our finding of undetectable EI
levels in CF patients homozygous for the AF508
mutation is consistent with other published
studies." The remaining three patients were
heterozygous for the AF508 mutation and had
E I levels within the reference range. These three
patients had symptoms suggestive of malabsorp-
tion and had subsequently received enzyme
supplements. On the basis of the EI concentra-
tion, enzymes were discontinued in all cases
without any deleterious clinical effect.
Of the seven patients judged to be pancreatic-
sufficient, five had E I levels within the reference
range and not statistically different from those
of the other patient groups. The remaining two
patients, aged 15 months and 14 years, did not
report symptoms suggesting malabsorption.
Both had 'normal' growth curves, negative stool
microscopy for fat globules, and vitamin A and
E levels within the reference range. Although
neither were receiving enzyme supplements, both
had undetectable levels of E 1. They have
subsequently been shown to have significant
steatorrhoea on analysis of 3-day faecal fat
collection (42 and 38 mmol/24 h, respectively;
reference range < 14 mmol/24 h). Enzyme ther-
apy has been commenced in the 15-month-old
child, resulting in growth from the 25th to above
the 50th centile for both height and weight.
These findings may be explained by the fact that
faecal fat microscopy might not be sufficiently
sensitive. Terbrack et al" reported a decline in
El levels in a CF child, from 100 Jig EI/g stool
to undetectable levels, over a period of 6
months. There was no information given about
the age, time of diagnosis or genotype of this
child. Another factor to be considered is that
some pancreatic-sufficient CF patients suffer
episodes of acute pancreatitis and this might be
associated with declining pancreatic function.
Ann Clin Biochem 1999: 36
CONCLUSION
The measurement of pancreatic El is a simple,
non-invasive method of assessing exocrine
pancreatic status in CF, requiring only a random
stool specimen. Currently, there is no informa-
tion on the potential dilutional effects of
diarrhoea. A study to evaluate dilutional
interference is planned, but pending further data
a formed stool sample should be used where
possible.
Elastase I measurements in this study have led
to useful changes in patient management and
appear to discriminate well between CF patients
with and without endogenous pancreatic func-
tion. We conclude that measurement of E I
appears to be a promising tool in the assessment
of pancreatic function in CF.
REFERENCES
Stern M, Geddes D. Cystic fibrosis: basic chemical
and cellular mechanisms. Br J Hosp MC!d 1996; 55:
237-40
2 Malfertheiner P, Buchler M, Muller A, Ditschuneit
H. Influence of extra-pancreatic digestive disorders
on the indirect pancreatic function test with
fluorescein dilauratc. Clin Physiol Biochem 1985; 3:
166--73
3 Mallory PA, Travis J. Human pancreatic enzymes:
Purification and characterization of a nonelastolytic
enzyme, protease E, resembling elastase. Biochem-
istry 1975; 14: 722-30
4 Largman C, Brodrick lW, Geokes C. Purification
and characterization of two human pancreatic
elastases. Biochemistry 1976: 15: 2491-500
5 Stein J, lung M, Sziegoleit A, Zeuzern S, Caspary
WF, Lembcke B. Immunoreactive elastase I.
Clinical evaluation of a new non-invasive test of
pancreatic function. Clin Chem 1996; 42: 222--{'
6 Soldan W, Henker J, Sprossig C. Sensitivity and
specificity of quantitative determination of pancrea-
tic elastase I in faeces of children. J Pediatr
Gastroenterol Nutr 1997; 24: 53-5
7 Loser C, Mollgaard A, Folsch U. Faecal elastase I:
a novel, highly sensitive, and specific tubeless
pancreatic function test. Gut 1996; 39: 580-6
8 Wallis C, Leung T. Cubitt D, Reynolds A. Stool
elastase as a diagnostic test for pancreatic function
in children with cystic fibrosis. Lancet 1997: 350:
1001
9 Terbrack HG, Gurtler KH, Huls G, Bittner-Dersch
P, Klor HU, Lindemann H. Human faecal elastase in
children. Monatsschr Kinderheiikd 1996; 144: 901--5
Acceptedfor publication 16 June 1999