Journal of Clinical Laboratory Analysis 5:410-414 (1991)

New Method for the Determination of Pancreatic Amylase Evaluated

Klaus Lorentz,’ Renze Baiq4 Peter M. B a ~ e r , ~ Herbert Keller,‘ Matti P ~ u k k a , ~ Juhani R U O S ~ ~ S U O , ~ Raija P ~ u k k a , ~ Sidney B. Rosalki,8 and Ying Foo,~ Dieter Seiler,2

Dietmar Nagel,2 Ulrich Naegele,3 and Peter Koller3 ‘lnstitut fur Klinische Chemie, Medizinische Universitat zu Lubeck, Lubeck, *KIinikum der Stadt Ludwigshafen,

lnstitut fur Klinische Chemie, Ludwigshafen and 3Boehringer Mannheim Gmb H, Evaluation Department, Mannheim, FRG; 4Division of Clinical Chemistry, Institute of Medical and Veterinary Science, Adelaide,

Australia; 5Wilhelminenspital der Stadt Wien, Zentrallabor, Vienna, Austria; ‘lnstitut fur Klinische Chemie und Hamatologie des Kantons St. Gallen, St. Gallen, Switzerland; 70ulun University Central Hospital, Laboratory,

Oulu, Finland; ‘Department of Chemical Pathology and Human Metabolism, Royal Free Hospital and School of Medicine, Hampstead, London, United Kingdom

We have carried out a multicenter evalua- tion of a new reagent carrier for Reflotronm’, specific for pancreatic amylase where the sali- vary isoenzyme is inhibited by two specific monoclonal antibodies. This new procedure combines easy handling with low imprecision (median CV < 3%) in control material, se- rum, heparinized blood, and plasma) and close correlation (r = 0.991 to 0.999) with estab- lished manual and automated methods. The same close correlation was found with values obtained from either venous or capillary finger-stick blood. Salivary amylase up to 54 kUiL (37°C) was inhibited to about 97%.

Endogenous interference by hemoglobin, bilirubin, triglycerides, cholesterol, or hema- tocrit was found to be negligible within a wide range of interferent concentrations. Out of a panel of 28 commonly used drugs it was shown that only two (ascorbic acid and par- acetamol), and then only at toxic concen- trations, caused a deviation in amylase activ- ity of greater than 10%. From the results of this study we conclude that this new method is suitable for highly precise and accurate measurements of pancreatic amylase in emer- gency and routine laboratories.

Key words: synthetic substrates, monoclonal antibodies, reflectance photometry, dry-reagent carrier

INTRODUCTION

Since pancreatic isoamylase has proved to be significantly more specific for the diagnosis of pancreatic diseases than total a-amylase (1 ,Ca-D-glucan glucanohydrolase, EC 3.2.1.1) (1 -3) a highly reliable technique for its estimation has been developed and evaluated (4,5).

This method exploits the synergistic action of the mono- clonal antibodies 66 C7 and 88 E8 (6) against salivary isoamylase. Most recently (7) this antibody technique has been applied for use with the ReflotronB system. The salivary com- ponent of amylase is bound by 66 C7 and inhibited by 88 E8 during its passage through the glass fibre layer of the reagent carrier.

We report on a multicentre evaluation of this new technique in several laboratories. In these evaluations special emphasis has been placed on the use of whole blood and on the com- plete removal of the salivary isoenzyme.

0 1991 Wiley-Liss, Inc.

MATERIALS AND METHODS

Procedure

The Reflotron@ pancreatic amylase dry-reagent carrier, the substrate indolyl-a-D-maltoheptaoside, and the ReflotronB sys- tem itself have been described elsewhere (7- 10). Sample mate- rials were 30 p l of capillary or venous blood, serum, or heparinized plasma.

Evaluation Sites

This study was the collaborative effort of seven clinical lab- oratories and the Evaluation Department of Boehringer Mann- heim GmbH.

Received May 23,1991; accepted August 8, 1991.

Address reprint requests to Klaus Lorentz, Institut fur Klinische Chemie, Medizinische Universitat zu Lubeck, Kronsforder Allee 71-73 D-2400 Lubeck, FRG.

TA

BL

E 1

. R

esul

ts of

the I

mpr

ecis

ion

Stud

y

Ref

lotro

n" p

ancr

eatic

am

ylas

e P-

amyl

ase

PNP

cv, %

A

ssig

ned

cv

, %

Ass

igne

d Sa

mpl

e mat

eria

l va

lue,

U/L

M

in.

Med

ian

Max

. va

lue,

U/L

M

in.

Med

ian

Max

.

With

in-s

erie

s im

prec

isio

n, n

= 1

0 re

plic

ates

per

seri

es, s

ix la

bora

torie

s C

ontro

l ser

a 1,

I8

serie

s 410

1.3

2.8

4.2

396

0.6

0.9

3.3

Con

trol s

era 2, 18

serie

s 640

1 .o

2.3

4.8

567

0.7

1.4

4.2

Hep

arin

blo

od, l

ow c

oncn

.

Hep

arin

blo

od, h

igh

conc

n.

Hep

arin

pla

sma,

low

con

cn.

Hep

arin

pia

sma,

hig

h co

ncn.

Day

-to-d

ay im

prec

isio

n (n

= 1

0)

9 se

ries

<I2

0 1.5

2.6

4.4

13 se

ries

>I20

1.5

2.3

3.1

9 se

ries

<I20

I .2

2.6

3.0

<I20

1 .o

3.0

5.1

I3 se

ries

>I20

2.1

2.8

4.4

>I20

0.6

I .4

3.5

Con

trol s

era

I, 6 se

ries

410

1.9

3.2

4.5

396

I .3

2.3

3.8

Con

trol s

era 2,6 s

erie

s 640

2.0

3.4

4.0

561

1.3

1.9

3.6

TA

BL

E 2.

Ref

lotr

onB

Pan

crea

tic A

myl

ase:

Met

hod

Com

pari

sons

for

Diff

eren

t Spe

cim

ens"

Met

hod

x,

Met

hod

y. A

ctiv

ity

Paire

d in

stru

men

t sa

mpl

e m

ater

ial

rang

e, U

IL

s a m

p 1 e s

Sl

ope

y-in

terc

ept

r

Uni

cam

SP

1800

Co

basm

Far

a E

PO

P 5060

Coba

sm B

io

Eppe

ndor

f 10

01 M

Co

basm

Bio

R

eflo

tron@

U

viko

n 860

Uvi

kon 860

Cob

as"

Bio

Blo

od

10-1

,800

Pl

asm

a 20- 1,600

Seru

m

20-2,650

Seru

m

20-2.200

Plas

ma

25-1,600

Seru

m

25-600

Cap

illar

y vs

. ven

ous b

lood

40-750

Cap

illar

y vs

. ven

ous p

lasm

a 40-750

Cap

illar

y bl

ood

vs. c

apill

ary

plas

ma

40-750

Urin

e 200-6,OOO

100

1 .00

4 10

0 1.

01

6.7

106

0.97

- 3.3

100

0.93

4.1

50

0.97

-3.0

98

0.98

2.6

60

0.99

1.3

60

0.93

10.6

185

I .04

68.8

60

1 .oo

- 2.4

0.999

0.997

0.996

0.998

0.995

0.991

0.998

0.999

0.998

0.992

= c

ompa

rison

met

hod;

y =

Ref

lotro

n" p

ancr

eatic

am

ylas

e.

41 2 Lorentz et al. $

f '

X

X

Fig. 1. Reflotronm pancreatic amylase: Comparison between capillary and venous blood samples.

The imprecision studies and method comparisons were undertaken in all laboratories whereas the urine experiments were carried out at one laboratory only. The isoenzyme study, the capillary blood and the interference experiments were per- formed at the Evaluation Department of Boehringer Mann- heim GmbH.

Comparison Methods

For all comparison experiments the pancreatic amylase test kit P-Amylase PNP (Boehringer Mannheim GmbH) was used according to the manufacturer's instructions either manually on photometer Unicam SP 1800, Eppendorf 1101M, and Kontron Uvikon 860 or automated with the Hitachi 737, Eppendorf EPOS 5060, Hoffmann-La Roche Cobas@Bio or CobasBFara.

Samples

For the within-series imprecision study the evaluators car- ried out collectively, 18 series of tenfold determinations for each sample, using ReflotronB control sera PrecinormB U and PrecipathB U as well as 22 series of tenfold determinations using fresh heparinized venous blood with either low or high

activities and, after prior splitting of the whole blood sam- ple, the corresponding plasma obtained after centrifugation.

The day-to-day imprecision was determined by measure- ment of the two control sera on 10 consecutive days.

The method comparison experiments were carried out with serum or fresh heparin-treated venous blood and plasma. Hep- arinized capillary blood was collected in MicrotainerB tubes (Becton Dickinson, Rutherford, NJ). After skin puncture the first drop of blood was discarded and then approximately 300 p1 of blood were taken.

The capacity for inhibition of salivary amylase was tested by adding increasing amounts of a serum with extremely high salivary amylase activity to a serum free of salivary amylase.

The studies on endogenous and drug interferences were performed using a procedure described elsewhere (1 1).

Statistical analysis of the method comparison data was per- formed following a nonparametric method described by Pass- ing and Bablok (12,13).

RESULTS AND DISCUSSION Imprecision

The intra-assay and inter-assay CVs of the reagent carrier method compare well with that found using the comparison

Determination of Pancreatic Amylase 413

TABLE 3. Reflotrona Pancreatic Amylase: Immunoinhibition in Relation to Salivary Amylase Activity

Salivary amylase Measured pancreatic added (U/L) amylase (UIL) Immunoinhibition (%)

Native serum 29 - + 1,038 57 97.3 + 5,300 I40 97.9 + 9,390 258 97.6 + 18,000 490 97.4 + 54,600 1347 97.6

TABLE 4. Reflotron@ Pancreatic Amylase: Endogenous Interferences

Interferent % recovery" Interferent % recovery"

Hemoglobin, g/L, serum

0 0.8 1.5 3.0 6.0

Bilirubin, FmollL, serum

5 37 75

180 250 3 20 360

Triglycerides, mmol/L, serum

0.7 3.3 5.7

10.7 16.0 23.4

100 97 96 96 96

100 99 96 97 96 99 96

100 100 101 100 101 98

Cholesterol, mmol/L, serum

2.6 4.5 6.5 8.3

10.5 12.5 14.8

Hematocrit, blood 0.3 0.4 0.45 0.5 0.55 0.6

Applied blood sample volume, +I

26 28 30 32 34

100 101 100 101 99

101 101

100 100 99

96-100 94-100 90-100

98 100 100 99 98

25.9 101

"Pancreatic amylase activity: 80- 120 U/L (37°C).

method, giving median CVs ranging from 2.3% to 3.4% com- pared to 0.9% to 3.0% for the P-amylase PNP method. The difference between intra-assay CV and inter-assay CV was surprisingly small (Table 1).

Method Comparison

Table 2 summarizes the results of the method comparison. The slopes of the regression equations vary from 0.93 to 1.04 and the intercepts are <10 U/L except for the urine samples. In this case the intercept was 68 U/L, but it has to be taken into account that these samples were diluted K + 3 with dis- tilled water. The activities (calculated) were up to 6,000 U/L thus leaving the intercept close to 1% of the highest activity measured.

TABLE 5. ReflotronB Pancreatic Amylase: Drug Interferences

Toxic concentrations, mdL"

1 Acetylsalicylic acid 2 Amphetamine 3 Ampicillin 4 Ascorbic acid 5 Bezafibrate 6 Caffeine 7 Carbochromen 8 Chloramphenicol 9 Ethaverine

10 Furosemide 11 Glibenclamide 12 Indomethacine 13 Methyldopa 14 Nicotinic acid 15 Nitrofurantoin 16 Oxytetracycline 17 Paracetamol 18 Phenprocoumon 19 Phenazopyridine 20 Phenytoin 2 1 Probenecide 22 Procaine 23 Pyridamol 24 Pyritinol 25 Quinidine 26 Sulfamethoxazole 27 Theophylline 28 Trimethoprim

I000 100

1000 300* 100 200

30 200

2 100

1 10 20

100 18

100 200* 40 25

100 lo00

10 20 20 60

600 100 20

'Concentrations with significant drug interference on the recovery in serum with Reflotrona pancreatic amylase are flagged.

The correlation coefficients ranging from 0.991 to 0.999 indicate close and good correlation between Reflotrona pan- creatic amylase and the comparison method, no matter which sample material is used for the Reflotron@ method.

Figure 1 shows the comparison of values obtained using venous blood and those obtained from capillary finger-stick blood from the same patients. It can be seen that there is no significant difference between the values, as indicated also by the regression equation y = 0 . 9 9 ~ + 1.3 (Table 3) and the correlation coefficient of 0.998.

lmmunoinhibition

The assessment of the immunoinhibition capacity of the anti- body coated reagent carrier is shown in Table 3. The data show that up to at least 50,000 U/L salivary amylase activity the immunoinhibition remains highly effective, close to 100%, and does not vary. The range is 97.3% to 97.9%. Neverthe- less, in rare cases of extremely high activity of salivary amy- lase (> 4,000 U/L) the remaining 2.5% activity may lead to false-positive results of Reflotrona pancreatic amylase.

Interferences

Table 4 lists the results of a study of endogenous interfer- ences on ReflotronB pancreatic amylase. Hemoglobin up to

41 4 Lorentz et al.

6.0 g/L (recovery 96%), bilirubin up to 360 kmol/L (96%), triglycerides up to 25.9 mmol/L (101%), and cholesterol up to 14.8 mmol/L (101%) did not show any effect on recovery compared to the comparison method values.

Hematocnt values between 0.30 and 0.45 did not affect the measurement of pancreatic amylase, and with values of 0.50 to 0.60, recovery rates of 96% to 90% are observed.

Although the recommended sample volume is 30 kl, van- ation in the applied volume of * 13% only gave a 2% differ- ence in recovery, indicating considerable insensitivity to volume changes for the test.

Out of 28 commonly used drugs (Table 5) only two, ascor- bic acid and paracetamol, interfered significantly, i.e., showed recovery of 125% and 85%, respectively. However, it has to be mentioned that the concentrations required were both at “toxic” levels.

Low imprecision, reliable method comparison, and effec- tive immunoinhibition in our opinion allow the conclusion that pancreatic amylase values determined with Reflotronm are highly specific, precise, and accurate not only for routine and emergency patient samples but also for samples with unusu- ally high salivary amylase contents. In addition, the interfer- ence experiments also support the good performance of the dry-reagent carrier chemistry.

REFERENCES

1. Melzi d’Eril GV, Bosoni T, Lesi C: Pancreatic amylase in serum for differential diagnosis of acute pancreatitis and acute abdominal diseases. Clin Chem 35:2142-2143, 1989.

2. Ventrucci M, Pezilli R, Gullo L, Plat6 L, Sprovieri G , Barbara L Role of serum pancreatic enyzme assays in diagnosis of pancreatic disease. Dig Dis Sci 34:39-45, 1989.

3 . Jensen DM, Royse VL, Bonello JN, Schaffner J : Use of amylase isoenzymes in laboratory evaluation of hyperamylasemia. Dig Dis Sci 32561-568, 1987.

4. Zaninotto M, Bertorelle R, Secchiero S , Plebani M, Burlina A: Assay of pancreatic amylase with use of monoclonal antibodies evaluated. Clin Chem 342552-2555,1988.

5 . Tietz NW, Burlina A, Gerhardt W, et al.: Multicenter evaluation of a specific pancreatic isoamylase assay based on a double monoclonal- antibody technique. Clin Chem 34:2096-2102, 1988.

6. Rauscher E, Gerber M: Pancreatic a-amylase assay employing the syn- ergism of two monoclonal antibodies. Clin Chim Actu 183:41-44, 1989.

7. Roth A, Wilk HE, Tritschler W: A new test in the Reflotronm system: pancreatic a-amylase. Clin Chem 34: 1289, 1988.

8. Steinhausen RL, Price CP: Principle and practice of dry chemistry sys- tems review. Recent Adv Clin Biochem 3:273-296, 1985.

9. Bais R, Badenoch J, Bayer PM, et al: a-Amylase determination with the Reflotrona reagent carrier system: use of whole blood, plasma, and serum, and effect of isoenzymes. Clin Chem 35:317-320, 1989.

10. Price CP, Koller PU: A multicentzr study of the new Reflotronm system for the measurement of urea, glucose, triacylglycerols, cholesterol, y-glutamyltransferase and hemoglobin. J Clin Chem Clin Biochem 26233-250, 1988.

11. Koller PU, Tritschler W, Carstensen CA: Interference studies on the Reflotron@ System. Lab Med 13399-402.1989.

12. Passing H, Bablok W: A new biometrical procedure for testing the equal- ity of measurements from two different analytical methods. Applica- tion of linear regression procedures for method comparison studies in clinical chemistry Part I . J Clin Chem Clin Biochem 21:709-720, 1983.

13. Passing H, Bablok W: Comparison of several regression procedures for method comparison studies and determination of sample sizes. Appli- cation of linear regression procedures for method comparison studies in clinical chemistry Part 11. J Clin Chem Clin Biochem 22:43 1 - 4 5 , 1984.