^^cnlvsg -

(

FISHERIES AND MARINE SERVICE

Translation Series No. 3998

Comparison of conventional methods of albumin, transferrin, andceruloplasmin determination in'serum with immunological'reference

methods

H..Schmidt, H. Ehelirig,

and D. Kraft

Original title: Vergleiche konventioneller Methbden zur Albumin-, Transferrin-,

`und Coeruloplasmin-Bestimmung im Serum mit immunologischenReferenzmethoden

From: Z. Klin. Chem. Klin. Biochem. 13(3): 117-121, 1975

Translated by the Translation Bureau (RL)

Multilingual Services Division

Department of the Secretary of State of Canada

pepartment of the Environment

Fisheries and-Marine Service

Halifax LaboratoryHalifax, N.S.

1977

20 pages typescript

Dil:PARTMENT OF THE SECRETARY OF STATE

TRANSLATION BUREAU

SECRÉTARIAT D'ÉTAT

BUREAU DES TRADUCTIONS

MULTILINGUAL SERVICES DIVISION DES SERVICES

Weee". wite CANADA

TRANSLATED FROM - TRADUCTION DE

German INTO •-• EN

English

DATE OF PUBLICATION DATE DE PUBLICATION

PAGE NUMBERS IN ORIGINAL NUMÉROS DES PAGES DANS

L'ORI GINAL

ISSUE NO. NUI•dRO

3

117-121 NUMBER OF TYPED PAGES

NOMBRE DE PAGES DACTYLOGRAPHIÉES

20 1975

YEAR

ANNÉE VOLUME

13

PUBLISHER -- ÉDITEUR

not shown

PLACE OF PUBLICATION LIEU DE PUBLICATION

--- not shown

Fii r 399cJP

DIVISION MULTILINGUES

AUTFIOR AUTEUR

H. SCHMIDT, H. EBELING, and D. KRAFT

TITLE IN ENGLISH •••• TITRE ANGLAIS

Comparison of Conventional Methods of Albumin, Transferrin, and Ceruloplasmin Determination in Serum with Immunological Reference Methods

TITLE IN FOREIGN LANGUAGE (TRANSLITERATE FOREI GN CHARACTERS)

• TITRE EN LANGUE ÉTRANGÉRE (TRANSCRIRE EN CARACTERES ROMAINS)

Vergleiche konventioneller Methoden zur Albumin-, Transferrin-, und Coeruloplasmin-Bestimmung im Serum mit immunologischen Referenzmethoden

REFERENCE IN FOREI GN LANGUAGE (NAME OF BOOK OR PUBLICATION) IN FULL. TRANSLI TERATE FOREIGN CHARACTERS. RÉFÉRENCE EN LANGUE ÉTRANGÉRE (NOM DU LIVRE OU PUBLICATION), Aà COMPLET, TRANSCRIRE EN CARACTÉRES ROMAINS.

Z. Klin. Chem. Klin. Biochem.

REFERENCE IN gNGLISH •-• RÉFÉRENCE EN ANGLAIS

Journal of Clinical Chemistry and Clinical Biochemistry

TRANSLATION BUREAU NO. 1089234 MINISTÈRE-CLIENT NOTRE DOSSIER N 0

Fisheries, Halifax Laboratory BRANCH OR DIVISION TRANSLATOR (IN(TIALS) Halifax, N.S. RL DIRECTION OU DIVISION TRADUCTEUR (INITIALES)

REQUESTING DEPARTMENT Environment

PERSON REQUESTING Dr. M. Kovacs DEMANDÉ PAR

YOUR NUMI3ER VOTRE DOSSIER N 0

07/03/77 DATE DE LA DEMANDE

UNEDITM TRANSLATION

For Information only

TRADUCTION NON RIEVISEE

Informniion setilomeni DATE OF REQUEST

29 1911

SOS.200.10.8 (R E V. 2/08)

71530-21-029•6333

DEPARTMENT OF THE SECRETARY OF STATE

TRANSLATION BUREAU

SECRÉTARIAT D'ÉTAT

BUREAU DES TRADUCTIONS

DIVISION DES SERVICES

MULTILINGUES

MULTILINGUAL SERVICES DiVISION

SOS-200-10-31

7830-21-025-I1332

-r-cr nit43q

CLIENT'S NO. DEPARTMENT DIVISION/BRANCH CITY

NO DU CLIENT MINISTi' RE DIVISION/DIRECTION VILLE

Fisheries • Environment . Halifax Laboratory Halifax, N.S

BUREAU NO. LANGUAGE . TRANSLATOR (INITIALS)

N° DU BUREAU LANGUE . TRADUCTEUR (INITIALES)

•

1089234 German RL Ain 20 1911

Z. Klin. Chem. Klin. Biochem., 13, no.3, 117-121 (1975)

Comparison of Conventional Methods of Albumin, Transferrin, p.117

and Ceruloplasmin Determination in Serum with Immunological

Reference Methods

H. Schmidt, H. Ebeling, and D. Kraft

From the Institut fUr Klinische Chemie und Klinische Bio-

chemie (Director: Prof. Dr. H.-J. Dulce) and the Medizinische

Klinik at the Klinikum Steglitz of the Free University of

Berlin

Received on January 13/February 21, 1975

Abstract

Radial immunodiffusion, which is a presently accepted reference method, was compared with the mechanized immunoprecipitation. Each of these specific methods was compared with two normal clinical-chemical methods for the determination of albumin and trans-ferrin, and an enzymic method for the determination of ceruloplasmin. The correlation coefficients, the mean slope for regression, and the percentage variation of the average values from those of the average values of the reference methods are presented. In the determination of albumin and ceruloplasmin, radial immunoditTusion and mechanized immunoprecipitation give the same results. Surprisingly, these two specific methods give different results in the determination of transferrin. In as far as there is no doubt concer-ning the accuracy of the mechanized immunoprecipitation, this method is recommended, since it also has the advantage of being less demanding in cost and time. . _

UNEDITED TRANSLATION For Inforrnalion ordy

TRADUCTION NON REVISEE Information souloment

2

With the introduction of specific immunological

methods in clinical chemistry, the question arises as to the

"accuracy" of conventional clinical-chemical analysis

methods.

Using albumin, transferrin or total iron-binding

capacity, and ceruloplasmin as examples, methods are being

compared and related to so-called reference methods.

The single radial immunodiffiü.sion method (ref. 1)

has so far been considered as the reference method of choice

for serum protein determinations. As a new reference method,

the mechanized specific-immunological determination employing

a nephelometric measuring technique (ref. 2) is compared with

the radial immunodiffusion method.

Material and Methods

From sera which were stored for approximately three

months at -20°C and which originated from chronically hemodi-

alized patients, analyses were carried out according to the

following methods:

Using the mechanized immunoprecipitation method, quan-

titative determinations of albumin, transferrin, and

3

ceruloplasmin were carried out nephelometrically according to

the procedure described by Ebeling (ref. 3). . Monospecific

antihuman sera supplied by BERINGWERKE AG, HYLAND, and

- DAKOPATTS were used.

2. Quantitative determinations of albumin, trans-

ferrin, and ceruloplasmin were also carried out on M-Partigen

immunodiffusion plates (BEHRINGWERKE AG) using the single

radial immunodiffusion method described by Mancini et al.

(ref. 1).

3. The total protein content of the serum was measured

with biuret reagent according to the method 'described by

Richterich (ref. 4) modified according to Hoffmann et al.

(ref. 5).

The protein electrophoreses were carried out on a

microscale on cellulose acetate film (SARTORIUS) in a BOS-

KAMP chamber, and analyzed after ponceau S coloration in an

Elphormat* (BENDER and HOBEIN). The albumin concentration

was calculated from the relative percentages of the albumin

fraction and the total protein content of the serum.

4. In addition, albumin was quantitatively determined

using the bromocresol green micro method according to

Schirardin et al. (ref. 6) as a colour test.

*Translator's note: Presumably a commercial designation.

4

5. The serum iron concentration (=Fe) was determined

with bathophenanthroline and the TECHNICON AutoAnalyzer in

accordance with the method described by Zak et al. (ref. 7).

6. The latent iron-binding capacity was determined

using the Irosorb-59 test (ABBOTT) based on ion exchange.

The radioactivity was measured ina well-type scintillation

counter (FRIESICKE and HOEPFNER).

7. The total iron-binding capacity of the serum was

determined manually using a micro method according to Ftthr

(refs. 8, 9).

8. As a clinical-chemical test for the ceruloplasmin p.118

determination, we used the enzyme test (DR. HAURY) which mea-

sures the oxidoreductase property of the ceruloplasmin via

the conversion of paraphenylenediamine.

For the immunological methods, protein standards and

control sera obtained from BERINGWERKE AG were used for

calibration and quality control respectively. Control sera

obtained from BEHRING[AIERKE AG in addition to sera obtained

from the DADE Company were also used for quality control of

the colorimetric methods and for the Irosorb method.

5

Results

The individual methods used were evaluated by determi-

ning, among other things, their precision. The different

methods were compared and their accuracy estimated by deter-

mining the mean regression lines and the percentage deviations

of the mean values from the mean values of the reference methods.

Table 1 shows the coefficients of variation of the

precision measurements using the various methods of determina-

tion. In the case of composite determinations, the coeffi-

cient of variation contains the relative error of propagation.

The concentrations of the precision controls give an

indication of the sensitivity of the particular method, and,

indirectly; allow one to infer the method-dependent serum

requirements.

In order to be able to compare total iron-binding

capacities with the results of the mechanized immunoprecipi-

tation and the radial immunodiffusion, we converted the trans-

ferrin concentrations into transferrin iron-binding capaci-

ties. In doing so, we assumed that at iron saturation each

transferrin molecule binds two iron atoms (=Fe +++ ) and

that the molecular weight of transferrin was 89000 (ref. 10).

Table 1: Precision of the methods of determination of a control serum.

1. 2.

Konzen- VK % tration der Prâ- zisions- kontrolle

Albumin g/1 . mechanisierte Immunprâzipitation 12 0,0188 0,66 +++

4. radiale Immundiffusion 60 3,27 0,66 ++ 5. Farbtest 20 42,0 2,28 ++ 6. Elektrophorese 5 42,9 2,20+++

7. Eisenbindungskapazitiit umo1/1 5. rnechanisierte Immunprâzipitation 15 4,315 2,16 +++ 4 • radiale Immundiffusion 28 78,85 1,00 ++ 8. latente Eisenbindungskapazitiit + Fe 8 59; 6 7,84 +++ 9. totale Eisenbindungskapazitât 20 74,7 4,19 ++

• 10. Coendoplasmin mg/1 or

U/1 5. mechanisierte « Immunpriizipitation 10 . 4,918 3,28 +++ 4 • radiale Immundiffusion 8 190,0 1,13 +i-

ll . enzymatisch 8 23,4 5,48+

+ in series ++ per series +++ relative propagated error per series

Translation of Numberea Items:

-- 1. concentration of precision control

2. coefficient of variation

3. mechanized immunoprecipitation

4 , radial immunodiffusion

5. colour test

6. electrophoresis

7. iron-binding capacity

8. latent iron-binding capacity + Fe

9. total iron-binding capacity

10. ceruloplasmin

11. enzymatically

Translator's note: Note that German "comma" = English decimal point.

7

The results of the -comparisonof the methods with one

another and with the reference methods are shown in Table 2.

The graphs show only the results of the comparisons

of the specific immunological methods. Apart from the experi-

mentally measured values and the mean regression lines, the

graphs also show the required or expected ideal relationships

between the methods.

Table 3 shows, among other things, the percentage

deviation of the mean value of a particular method from the

mean value of the corresponding reference method.

Discussion

In order for the reference methods used here to be

regarded as accurate, the antigen-antibody specificity and

quality of the analysis methods should be optimal. Moreover,

stable, qualitatively sufficiently concentrated antisera

should be available, as well as protein calibration standards

of extremely high quality, stability, and weight precision,

especially in the case of high-precision quantitative analy-

ses. The accuracy of the methods is determined by the extent

to which these conditions are satisfied.

p.119

8

Table 2: Comparison of inethods by means of the correlation coefficientand the mean regression line.

1.

N r

AlbuminRID/AiP 39 0,935

2. Farbtcst/AIP 42 0,937Elektrophorese/AIP 39 0,9131'arbtest/RID 42 0,975

5. Elektrophorese/RID 45 0,954

6.l arbtest/ 45 0,946Elektrophoresc

7. EisenbinduutgskapazitâtAIP/R1D 40 0,921latent f•.13K + Fe/AIP 37 0,907total EBK/AIP 35 0,938latent 1:BK + Fe/RID 34 0,933total EBK/RID 40 0,944latent EBK+Fc/ 38 0,948total- E$K

8. Coeniloplasmin

AIP/RID 37 0,9239. enzymatisch/AIP 37 0,915

10. ; cnzymatiscli/RID 37 0,882

mittlereRegressionsgerade(y=axtb)

a b

1,11 - 3,641,08 - 5,561,11 - 5,330,98 - 1,491,02 - 1,981,04 + 0,30

0,95 + 8,751,12 - 18,020,84 = 3,821,30 - 20,971,09 - 13,291,13 + 0,63

0,99 + 9,9615,43 - 90,30

1,20 - 0,70 ++++13,14 - 15,46

1,03 - 0,20++++

tt+I in this case, the measured conversion factor of 12.82was used (see text).

RID = radial immunodiffusion

AIP = mechanized immunoprecipitation

latent EBK + Fe = latent iron-binding capacity + Fe

total EBK = total iron-binding capacity

Translation of Numbered Items:

1. mean regression line (y = ax ± b) 9. enzymatically/AIP

2. colour test/AIP 10. enzymatically/RID

3. electrophoresis/AIP

11. colour test/RID

5. electrophoresis/RID

6. colour test/electrophoresis

7. iron-binding capacity

8. ceruloplasmin

Table 3: Percentage deviations of the mean value from the mean value of the reference method.

1.

N Mittelwerts- abweichung in %

WI • 2 . mechanisierte

Immunpriizipitation 40 36,3 3. radiale Imnuindiffusion 42 36,2 — 0,25 4. rubtest 43 38,8 + 7,00 5, Elektrophorese 43 37,9 + 4,50

3. radiale lmmundiffusion 42 36,2 2. rnechnisierte

Immunprâzipitation 40 36,3 + 0,25 14. Farbtest 43 38,8 + 7,00 5. Elektrophorese 43 37,9 + 4,50

6. Eisenbindungskapazitat prnol/I

2. mechanisierte Immunprâzipitation 38 54,8

3. radiale Immundiffusion 38 62,7 + 14,40 7. latente

8. : tEoi tsae lnebEinisdetikapazitât

nnbgis + Fe 37 63,6 + 16,10

ndungskapazitât 38 70,9 + 29:40

• radiale Immundiffusion 38 62,7

2 . mechanisierte • Immunprâzipitation 38 54,8 — 12,60

7 , latente ' • Eisenbindungskapazitât + Fe 37 63,6 + 1,40 8. • totale Eisenbindungskapazitât 38 70,9 + 13,10

9. Coeruloplasmin mg/1 or U/t

2. meehanisierte :Immunprâzipitation 37 349,80

3. ' radiale Immundiffusion 37 354,96 + 1,48 10. enzymatisch 37 27,73 + 1,65 ++44

3. radiale Immundiffusion 37 354,96

2 . tnechanisierte • Imnninpriizipitation 37 349,80 — 1,45

10. enzymatisch 37 27,73 + 0,14 ++++

+++A- in this case, the measured conversion factor of 12.82 was used (see text).

9

Affiumin

(for translation of numbered items, see next page)

10

Translation of Numbered Items in Table 3:

1. mean value deviation in %

2. mechanized immunoprecipitation

3. radial immunodiffusion

4. colour test

5. electrophoresis

6. iron-binding capacity

7. latent iron-binding capacity + Fe

8. total iron-binding capacity

9. ceruloplasmin

10. enzymatically

1 1

It is evident from the comparison of the albumin

methods, that the single radial immunodiffusion method and the

mechanized nephelometric immunoprecipitation method yield the

same results. Thus, at equal precision, the two specific

methods of albumin determination can be regarded as equivalent

reference methods. However, since the mechanized immunopre-

cipitation method is less demanding in cost and time, we con-

sider it today as the preferable method (refs. 3, 11). By

comparison, the two unspecific methods of albumin determin-

tion, the colour test with bromocresol green, and the electro-

phoretic separation and quantification of the relative percen-

tages through total protein determinations yield values which

are too high. Evidently, the dyeability of albumin after the

serum electrophoresis with ponceau S is greater than that of

other protein fractions. This is no doubt due, among other

things, to the fact that the so-called albumin band does not

represent a protein-chemically homogeneous fraction, which

means that the measurement includes also other proteins, apart

from the difference in the protein-dye affinity (ref. 12).

Nor is the bromocresol green method only specific to albumin,

because other proteins evidently also take part in the reac-

tion and thus give rise to higher apparent albumin concentra-

tions (ref. 11). However, the two latter methods correlate

well, as was confirmed by Schirardin et al. (ref. 6) and

Booij (ref. 12).

12

In order to be able to compare the two specific immu-

nological methods of transferrin determination with the unspe-

cific methods, i.e. the serum iron-binding capacity methods,

the transferrin concentration was expressed as transferrin

iron-binding capacity in accordance with Haeckel et al. (ref.

13). As was reconfirmed by our results, the concentrations

obtained with the unspecific total serum iron-binding capacity

methods, the Irosorb-59 plus serum iron method, and, above all,

the method according to Führ, are higher than the concentra-

tions obtained with the specific methods. The fact that

there exists a discrepancy between the total and the transfer-

rin iron-binding capacity was observed by Stojceski et al.

(ref. 14) and Scuro et al. (ref. 15), among others. Van der

Heul et al. (ref. 16) were able to show that other iron-

binding serum components also enter into the total serum iron-

binding capacity apart from transferrin. For this reason

alone, a determination of the total iron-binding capacity can

no longer be equated with a transferrin determination (refs.

13, 17).

Contrary to our own expectations and those of other

authors (ref. 13), a comparison of the two specific methods

shows that the reference method for transferrin commonly used

so far, i.e. single radial immunodiffusion, yields higher

concentrations than the mechanized precipitation method. The

13

mean value obtained by immunodiffusion was approximately 14%

higher than the value obtained by nephelometry. This differ-

ence is statistically significant at the p < 0.025 level using

the t-test.

So far, we have been unable to resolve the question

as to the causes of this astonishing discrepancy between the

two methods. It would be interesting to investigate whether

this discrepancy in the results is due to physicochemical dif-

ferences in the principles of measurement or whether, as a

result of different, higher concentration ratios than in the

nephelometric method, the measurement by the radial immunodif-

fusion method includes an "unspecific"'component in the

antigen-antibody reaction and thus leads to higher values.

It should be pointed out that the same antigen standards were p.120

used in both methods, and that the quality and specificity

of the Partigen-plate antibodies can probably be assumed to -

correspond to the quality and specificity of the antisera in

the bottles from the same Company.

As long as the question as to which one of the two

u specific" methods yields the correct result remains unresol-

ved, it is difficult to accord one of the two methods the

rating of "reference" method.

14

It is evident from a comparison of the ceruloplasmin

methods that the concentrations determined via radial immuno-

diffusion and nephelometric immunoprecipitation differ only

slightly. Thus, it can be concluded that these two methods

are equivalent reference methods, as in the case of albumin.

The values obtained in the present study by oxidoreduc-

tase activity determination of ceruloplasmin correlated well

with the values obtained with the reference methods; however,

in contrast to comparisons of mg/1 to U/1 such as 10.4/1 and

10/1 reported in the literature (refs. 18, 19), we obtained

values of mg to U such as 12.82 to 1. As in the study by

Haralambie (ref. 19), our ceruloplasmin-protein standard also

yielded lower enzyme-kinetically determined values than

immunologically determined values, assuming a ratio of 10/1

mg/1 immunologically determined to U/1 enzyme-kinetically de-

termined. The good agreement of 10/1 of immunologically and

enzyme-kinetically determined values was obtained by Haralambie

only if fresh sera were used for the enzyme-kinetic method

(ref. 19). Despite the enzyme activity losses described (cf.

loc. cit. [ref. 19]), the immunological activity of the ceru-

loplasmin remains preserved (ref. 19). The sera used in our

determination were all frozen for approximately three months

at - 20°C, and yielded approximately the same activity differ-

ence immunologically to enzymatically as the lyophilized

•

- */*

l . o

1'7 10 10

1 30 40 50 60

Albumin (radiale (mmundit fusion) (gill

bO

15

0

cd 4-)

a) • 60

o --°2

50

rtj

H ce40

I1) E N e

g 30

' Ic5 0 6 (1) /

albumin (radial immunodiffusion) [g/l]

Fig. 1: Comparison of methods: albumin

mean regression line

ideal curve.

75

50

25

o I

Transferrin (mechamsierte Immunpr5zipitation) [/cmotllj

i i25 50 75 100 125

transferrin (mechanized immunoprecipitation) [umol/1]

Fig.-2: Comparison of methods: transferrin. The concentrations areexpressed as transferrin iron-binding capacities.

16

.j

mean regression line

------- ideal curve

bD

•

0 0 •

•i--1 700 — .

/ W .

0 /

rci • ..:-.: 0 • cn : ./.1.

. e . 500 H (2, •d

H a .3 'a •

r—I g 400 — . • • 5," 0 E • .7 e • • .1-1 E rd „„ • •

• •s"

O — ••e7r • e $-1 i0 ; 30 - a

■-•• 0 -1:: • • • -ç-

.40////// •

..-1 5 200 — 0 11 • W g . at3 -5 H ts-, Pit • 13 100 - 0 .

° a-

C) O 100 200 300 400 500 600 Coeruloptasmin (mechanisierte immunpnizipitation) Img/11

ceruloplasmin (mechanized immunoprecipitation) [mg/l]

Fig. 3: Comparison of methods: ceruloplasmin

mean regression line

ideal curve

17

18

ceruloplasmin standard of BEHRINGWERKE AG with 300 mg/1 to

23.4 U/1 (see Table 3) .

The ratio, which was obtained from the ceruloplasmin

standardized on a weight basis to the enzyme activity measured

by us, was used in Table 3 as conversion factor for the per-

centage deviation of the mean values of the enzymatic method

from the mean values of the reference methods.

Since there is evidently no immunological activity

loss to be expected in the two specific immunological methods,

in contrast to the enzyme-kinetic ceruloplasmin determination,

immunological values are more reliable and reproducible.

Measurements obtained by specific quantitative ana-

lysis methods provide physicians with reliable information.

Therefore, the unspecific methods still widely used in cli-

nical chemistry should be replaced by more specific methods.

Provided that there is no doubt about the "accuracy"

of the methods, we recommend the mechanized immunoprecipita-

tion for serum protein determinations. If it may be assumed

that, as part of the trend towards general automation, clini-

cal laboratories have at their disposal the necessary facili-

ties, including continuous-flow fluoronephelometers (TURNER or

I

19

TECHNICON), the method of mechanized immunoprecipitation is

superior to immunodiffusion for reasons of reagent costs and

speed, and, as far as the reproducibility of the results is

concerned, it is not inferior to immunodiffusion.

Acknowledgements: The authors are indebted to ABBOTT,

BEHRINGWERKE AG, DAKOPATTS, and HYLAND for providing test

packs.

Bibliography

1. Mancini, G., Carbonara, A. O. & Heremans, J. F. (1963), Immunochemistry 2, 235-254.

2. Eckman, I., Robbins, J. B., Van Den Hamer, C. J. A., Lentz, J. & Scheinberg, I. H. (1970), Clin. Chem. 16, 558— 561.

3. Ebeling, H. (1973), these Z. 11, 209-214. 4. Richterich, R. (1971), Klinischc Chemie, Theorie und Praxis,

3. Aufl. S. 305-309, S. Karger, Basci. 5. Hoffmann, J. P. & Richterich, R. (1970), diese Z. 8, 595—

598. 6. Schirardin, H. & Ney, J. (1972), diese Z. 10, 336-343. 7. Zak, B. & Epstein, E. (1965), Clin. Chem. 11, 641-644. 8. Führ, J. (1965), Med. Monatsschrift, 19, 281-283. 9. Führ, J. & Stary, E. (1970), Technicon Symposium, Sonder-

druck R 874. 10. Bambach, M. N. (1966), Klin. Wochenschr. 44, 1276—

1284.

11. Wegfahrt, P. F., Fish, M. B., Aldana, F. B. & Aronson, S. B. (1971), Technieon International Congress 1970, Vol. 1, Clinical, S. 21-24, Thurman Associates, USA.

12. I3ooij, J. (1972), Clin. Chim. Acta 38, 355-362. • 13. Haeckel, R., Haindl, Fi. , Hultsch, E., Mariss, P. & Oellerich,

M. (1973), diese Z. 11, 529 -534. 14. Stojceski, T. K., Malpas, J. S. & Witts, L. J. (1965), J. Clin.

Pathol. 18, 446-452. 15. Scuro, L. A., Dobrilla, G., Lo Cascio, V., Bosello, O.,

D'Andrea, F. & lnnecco, A. (1972), Acta Hep.-Gastroenterol. 19, 90-98.

16. Van Der lieu', C., Van Eijk, H. G., Wiltink, W. F. & Leijnsc, B. (1972), Clin. Chim. Acta 3,9, 347-353.

17. Koch, C.D. & Ritter, U. (1974), Diagnostik 7, 766-770. 18. Jérôme, H. & Girault, M. (1969), Ann. Biol, Clin., 27,

371-385. 19. Haralambie, G. (1969), diese Z. 7, 352-355.

20

Author's Address:

Dr. licrwig l:bclingInstitut für Klinischc Chemieund Klinischc BiochemicKlinikum Stcglitzder f'rcicn Universit5t Berlin

. 1 Berlin 45,Hindenburgdamm 30

Translation of German Bibliographical Entry:

4. Clinical chemistry, theory and piactice.