323
Clinica Chimica Acta, 92 (1979) 323-328 Biomedical Press 0 El sevier/North-Holland
CCA 1002
DETERMINATION AN AUTOMATED
OF PREGNANCY ZONE PROTEIN IN SERUM USING IMMUNOPRECIPITIN REACTION METHOD
0. BERG * and L. HEMMINGSEN Department (Received
of Clinical Chemistry, Central Hospital, Nykbbing Falster (Denmark) October
30th,
1978)
Summary Determination of pregnancy zone protein in serum by means of an automated immunoprecipitin reaction method and pretreatment with polyethylene glycol for reducing the serum blanks is described. Using this procedure the sensitivity of the method was greatly improved, from originally 100 mg/l to 1 mg/l. The precision “between days” was 9.9%. A positive and significant correlation to electroimmuno assay is demonstrated. In apparently healthy controls the median serum pregnancy zone protein in females was 38 mg/l (2-91 mg/l) and in males 2 mg/l (O-20 mg/l). No correlation between serum pregnancy zone protein and age could be demonstrated. In malignant diseases our results seem to confirm a relationship between increased pregnancy zone protein and spreading of the tumours. Serum pregnancy zone protein in disseminated malignancies are increased significantly compared to controls. In males 75% with disseminated tumours have elevated serum pregnancy zone protein.
Introduction Pregnancy zone protein (PZP) was first described by Smithies in 1959 using starch gel electrophoresis on sera from pregnant women [ 11. Later it has been shown by different methods that the PZP concentration of non-pregnant healthy women is less than 100 mg/l 12-41, in healthy men less than 30 mg/l, usually only trace amounts [ 2-41. Beside pregnant women [ 2,5--81 PZP values are elevated in oestrogen treated women [3,5,9] and in men treated with diethylstilboestrol [2,5,7]. Until now the biological role and the clinical value of PZP has not been clear. Immunosuppressive properties of PZP have been
* To whom correspondence
should be addressed.
321
demonstrated [lO,ll] and other reports have shown increasing amounts of PZP in sera of patients with various malignancies [ 2,5,12-151. The determination of specific proteins in low concentrations in serum samples (below 100 mg/l) is normally not possible using the standard method of the Technicon Auto-Analyzer (AIP system) because of the high serum blanks. Recently Hellsing and Engstrijm [16] described a pre-treatment procedure of serum samples for the AIP system using polyethylene glycol for reducing the serum blanks. On this basis we have improved the automated immunoprecipitin reaction and stabilized the process by adding bovine serum albumin to the antibody solutions. The assay is compared with electroimmuno assay (EIA). Besides presentation of the method, this paper deals with a preliminary clinical study of PZP concentrations in normal healthy subjects and in various stages of different malignancies. Methods Automated immunoprecipitin reaction method (AIP assay) Standards: As a primary standard we used SP-3 Standard (Behring Institut, Marburg, F.R.G.) of 1600 mg/l. As a secondary standard we used a serum pool from pregnant women of 800 mg/l. Pretreatment of standards and serum samples: 1 ml serum is mixed with 1 ml pretreatment-R (20% polyethylene glycol mol. wt. 6000 in NaCl 0.9%), after 30 min the mixture is centrifuged and the supernatant is filtered through a Millipore filter 0.22 I_rm. The standards are diluted (1 + 79) and (1 + 319) with predilution-R (NaCl 0.9%), while the serum samples normally are diluted (1 + 1). Antiserum: Monospecific antibody to human pregnancy zone protein PZ (SP-3) (Dacopatts, Copenhagen, Denmark) is pretreated and diluted with antiserum-diluent (10% polyethylene glycol mol. wt. 6000 and 0.3% Bovine Albumin in NaCl 0.9%) in the proportion 1 : 40. After 30 min the mixture is centrifuged and the supernatant is filtered through a Millipore filter 0.22 pm. Nephelometry: The automated precipitin reaction method is carried out on the Technicon Auto-Analyzer for specific proteins. The predilution circuit is excluded in favour of manual predilution. The pretreated samples are aspirated at the rate of 50 pl/min and the antiserum-dilutions are aspirated at the rate of 1.00 ml/min. The reaction time is 6 min and the sample speed is 60 samples per hour. The sample/wash ratio is 1 : 1. The sample blanks are run using the antiserum-diluent instead of the diluted antiserum. The peak heights of the blanks and the specific reactions are stored on magnetic tape via an interface and calculator for later calculation, using the coordinates of the two standards (high and low) and the power model (y = a. xb) which in a double logarithmic coordinate system gives a straight line relationship (In y = b In a). The calculator (Hewlett Packard 9815 A) corrects automatically for base-line drift, and on the basis of the stored data calculates the sample values from the peak height differences. Precision: The precision of the method is “Within-day” C.V.% = 3.6% and “between-day” C.V.% = 9.9%. Accuracy: The accuracy is determined by means of comparison studies and
325
Fig. 1. The correlation of pregnancy zone protein methods, rocket electrophoresis and nephelometry.
(PZP)
concentration
measured
by two immunochemical
recovery studies. The recovery studies include 20 different patient samples analysed before and after adding of 71 mg/l of PZP from a reference serum. On an average the recovery was 86 mg/l (121%) with an observed range of 62-135 mg/l(72%-157%). Electra immunoassay Rocket electrophoresis is carried out by the method of Laurel1 agarose gel in sodium barbital pH 8.6, 0.017 M with an antiserum tion of 2%. The standard (serum pool 800 mg/l) is used undiluted dard) and diluted 1 + 9 (low standard). The serum samples are treated and undiluted 5 ~1 of each. The electrophoresis is run for Volt. The precipitates are stained with Ponceau S.
[17] in 1% concentra(high stanapplied un16 h at 25
Correlation studies As demonstrated in Fig. 1 there is a positive (b = 0.81) and significant 0.92, p < 0.001) correlation between the AIP method and the EIA method.
(r =
Material The controls (C) consisted of 111 healthy blood donors, 56 women aged 19-60 years and 55 men aged 19-64 years. The patient material included 28 women aged 22-87 years and 32 men aged 35-81 years, suffering from various malignancies. The women mainly had cancer of the breast, lungs or the female genital system. Among the men’s cancer of the lungs and the gastrointestinal tract were predominant. The diagnoses were confirmed by biopsy, during operation or at autopsy. The malignancies were graduated into three stages. Stage I (N,M,): Patients with localized tumours but without regional lymph node metastases or distant metastases. Stage II (NiM,-,): Patients with histologically verified lymph node metastases but without distant metastases. Stage III (NiMi): Patients with distant metastases. Stage I included 7 women
326
and 7 men, stage II 7 women and 10 men and stage III 14 women and 15 men. None of the women were pregnant and nor were any of the patients being treated with oestrogens. For statistical evaluations the non-parametric Mann-Whitney test was employed. We consider p values CO.01 as significant, 0.01 < p < 0.05 as possibly significant, and p > 0.05 as insignificant. Results The PZP values in sera from healthy subjects and in the three different stages of various malignancies are shown in Table I and in Fig. 2. Women: The interval of the PZP values in the three stages of malignancy is almost identical with the 0.95 reference interval of the controls. In stage I all observations are within the 0.95 reference interval, while 2 out of 7 in stage II and 3 out of 14 in stage III are above the 0.95 reference interval. On the other hand 8 out of 56 healthy women have PZP values below the interval of women with malignant disease.
RZ .P. mg/l
.
Female
. .lOO
. :
.
.
;
. __-.---
.
.
68
1 C f?Z.P.
I
__&__ . .
N*M,
N M.
NM,
Male
mg/f 100
Fig. 2. Serum PZP values of healthy male and female controls (C) related to patients with various malig. nancies. Localized tumour is indicated by NoMo. tumour with regional lymph node metastases by NlMo and disseminated tumour by N1 Ml. The median values are indicated by horizontal dot-and-dash lines.
321
TABLE I THE AVERAGE CONCENTRATION (MEDIAN) AND INTERVAL DONORS (CONTROLS) AND IN PATIENTS WITH DIFFERENT
OF PZP IN HEALTHY STAGES OF VARIOUS
BLOOD MALIG-
NANCIES Reference interval. 0.95. Females n
Controls (C) Stage I (NoMo). localized tumours Stage II (N, MO), regional lymph node metastases Stage III (N1 Ml ). distant metastases
Males Median
Interval
(mgP)
cmgn,
56 7 7
38 34
a4
2-91 19-87 21-132
14
86
13-99
n
Median
Interval
(mg/I)
(mgP)
55 7 10
2 11 12
O-20 O-17 o-43
15
47
o-93
Table II shows the significant differences of the PZP values between the different stages. The PZP values in patients with stage II and stage III tumours are increased significantly compared with the controls, while possible significance is demonstrated versus stage I tumours. Men: All PZP values in stage I are within the 0.95 reference interval. In stage II 3 out of 10 and in stage III 11 out of 15 are above the 0.95 reference interval. As demonstrated in Table II the PZP values in stage III tumours are significantly increased compared with controls, stage I and stage II tumours as well. Possible significance is demonstrated between the controls and stage II tumours. No correlation between PZP values and age was demonstrated either in the female or in the male material. Discussion In accordance with other investigations [2,4,14] we find higher PZP values in serum from non-pregnant, non-oestrogen treated women (38 mg/l) compared with healthy men (2 mg/l). The PZP values in cancer patients reported in previous publications have TABLE II THE SIGNIFICANT DIFFERENCES OF THE PZP VALUES CIES COMPARED INDIVIDUALLY AND WITH CONTROLS No significant difference is indicated by -,
I vs. controls II vs. stage I II vs. controls III vs. stage II III vs. stage I III vs. controls
Females
Males
-
-
(+I + (+) +
STAGES
OF MALIGNAN-
possibly significant by (+) and significant by +.
Statistical significance
Stage Stage Stage Stage Stage Stage
IN VARIOUS
(+) + + +
_
328
been contradictory. Some authors have demonstrated elevated values in both sexes, specially in cancers with distant metastases [5,13,15]. Berne [14] reported, that females frequently develop detectable elevation of PZP values in cancer, but males do not. Finally, some investigators have been unable to demonstrate elevated PZP values in cancer patients at all. Our results seem to confirm a relationship between increase in the PZP values and spreading of the malignant disease in both sexes. The diagnostic value of a single PZP determination concerning malignancy contra no malignancy and in differentiation between various tumour stages appears limited. We found the highest diagnostic value in men with distant metastases. PZP values higher than 20 mg/l indicate distant metastases. On the other hand serum PZP values below 20 mg/l in men do not exclude distant metastases. In women PZP values higher than 100 mg/l indicate a malignant disease with either regional lymph node metastases or distant metastases, while PZP values below 10 mg/l seem to exclude tumours. In conclusion our results demonstrate: (1) In cancer detection the diagnostic value of a single PZP determination is limited. (2) A disseminated malignant disease must be suspected in males with a PZP value higher than 20 mg/l. (3) Longitudinal studies of the PZP values may be of considerable value in monitoring cancer patients or as a residue control. Serial PZP determinations in patients with different malignancies in various tumour stages are under investigation. References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Smithies, 0. (1959) Adv. Protein Chem. 14,65 Bohn, H. (1974) Arch. Gyniikol. 217.219-231 Stimson, W.H. and Sinclair. S.M. (1974) FEBS L&t. 47,190 Von Schoultz, B., Damher. M.-G. and St&brand. T. (1976) Protides of Biological Fluids (Peeters. H., ed.). Vol. 24, pp. 189-192, Pergamon Press, Oxford Home, C.H.W., McLay, A.L;C., Tavadia. H.B.. Carmichael, I., MaIlinson, A.C., Young Laiwah, A.A.C., Thomas, M.A. and MacSween, R.N.M. (1973) CIii. Exp. Immunol. 13,603--611 Afonso, J.F. and de Alvarez. R.R. (1963) Am. J. Obstet. Gynecol. 86,815 Cooper, D.W. (1963) Nature 200.892 Beckman, L.. Von Schoultz. B. and Stigbrand, T. (1973) Acta Obstet. Gynecol. Stand. 51. 157-160 Beckman, L.. Von Schoultz. B. and Stigbrand. T. (1971) Acta Obstet. Gynecol. Stand. 50, 369371 Hagen. C. and Fr#land. A. (1972) Lancet i. 1185 (Letter to the editor) Von Schoultz, B., Stigbrand. T. and Tiimvik. A. (1973) FEBS Lett. 38.23 Than, G.N., Csaba, F., Szab6, D.G.. Karg, N.J. and NovPk, P.F. (1975) Arch. Gynlkol. 218, 125130 Stimson. W.H. (1975) J. CIin. Pathol. 28.868-871 Berne. B.H. (1976) Protides of Biological Fluids (Peeters. H.. ed.). Vol. 24, PP. 165-179. Pergamon Press. Oxford Anderson, J.M.. Stimson, W.H. and Kelly. F. (1976) Br. J. Surg. 63.819-822 HeUsing. K. and EngstrBm, H. (1977) Stand. J. Clin. Lab. Invest. 37, 529-536 LaureII, C.-B. (1972) Stand. J. CIii. Lab. Invest. 29, Suppl. 124,21-37
Clinica Chimica Acta, 92 (1979) 323-328 Biomedical Press 0 El sevier/North-Holland
CCA 1002
DETERMINATION AN AUTOMATED
OF PREGNANCY ZONE PROTEIN IN SERUM USING IMMUNOPRECIPITIN REACTION METHOD
0. BERG * and L. HEMMINGSEN Department (Received
of Clinical Chemistry, Central Hospital, Nykbbing Falster (Denmark) October
30th,
1978)
Summary Determination of pregnancy zone protein in serum by means of an automated immunoprecipitin reaction method and pretreatment with polyethylene glycol for reducing the serum blanks is described. Using this procedure the sensitivity of the method was greatly improved, from originally 100 mg/l to 1 mg/l. The precision “between days” was 9.9%. A positive and significant correlation to electroimmuno assay is demonstrated. In apparently healthy controls the median serum pregnancy zone protein in females was 38 mg/l (2-91 mg/l) and in males 2 mg/l (O-20 mg/l). No correlation between serum pregnancy zone protein and age could be demonstrated. In malignant diseases our results seem to confirm a relationship between increased pregnancy zone protein and spreading of the tumours. Serum pregnancy zone protein in disseminated malignancies are increased significantly compared to controls. In males 75% with disseminated tumours have elevated serum pregnancy zone protein.
Introduction Pregnancy zone protein (PZP) was first described by Smithies in 1959 using starch gel electrophoresis on sera from pregnant women [ 11. Later it has been shown by different methods that the PZP concentration of non-pregnant healthy women is less than 100 mg/l 12-41, in healthy men less than 30 mg/l, usually only trace amounts [ 2-41. Beside pregnant women [ 2,5--81 PZP values are elevated in oestrogen treated women [3,5,9] and in men treated with diethylstilboestrol [2,5,7]. Until now the biological role and the clinical value of PZP has not been clear. Immunosuppressive properties of PZP have been
* To whom correspondence
should be addressed.
321
demonstrated [lO,ll] and other reports have shown increasing amounts of PZP in sera of patients with various malignancies [ 2,5,12-151. The determination of specific proteins in low concentrations in serum samples (below 100 mg/l) is normally not possible using the standard method of the Technicon Auto-Analyzer (AIP system) because of the high serum blanks. Recently Hellsing and Engstrijm [16] described a pre-treatment procedure of serum samples for the AIP system using polyethylene glycol for reducing the serum blanks. On this basis we have improved the automated immunoprecipitin reaction and stabilized the process by adding bovine serum albumin to the antibody solutions. The assay is compared with electroimmuno assay (EIA). Besides presentation of the method, this paper deals with a preliminary clinical study of PZP concentrations in normal healthy subjects and in various stages of different malignancies. Methods Automated immunoprecipitin reaction method (AIP assay) Standards: As a primary standard we used SP-3 Standard (Behring Institut, Marburg, F.R.G.) of 1600 mg/l. As a secondary standard we used a serum pool from pregnant women of 800 mg/l. Pretreatment of standards and serum samples: 1 ml serum is mixed with 1 ml pretreatment-R (20% polyethylene glycol mol. wt. 6000 in NaCl 0.9%), after 30 min the mixture is centrifuged and the supernatant is filtered through a Millipore filter 0.22 I_rm. The standards are diluted (1 + 79) and (1 + 319) with predilution-R (NaCl 0.9%), while the serum samples normally are diluted (1 + 1). Antiserum: Monospecific antibody to human pregnancy zone protein PZ (SP-3) (Dacopatts, Copenhagen, Denmark) is pretreated and diluted with antiserum-diluent (10% polyethylene glycol mol. wt. 6000 and 0.3% Bovine Albumin in NaCl 0.9%) in the proportion 1 : 40. After 30 min the mixture is centrifuged and the supernatant is filtered through a Millipore filter 0.22 pm. Nephelometry: The automated precipitin reaction method is carried out on the Technicon Auto-Analyzer for specific proteins. The predilution circuit is excluded in favour of manual predilution. The pretreated samples are aspirated at the rate of 50 pl/min and the antiserum-dilutions are aspirated at the rate of 1.00 ml/min. The reaction time is 6 min and the sample speed is 60 samples per hour. The sample/wash ratio is 1 : 1. The sample blanks are run using the antiserum-diluent instead of the diluted antiserum. The peak heights of the blanks and the specific reactions are stored on magnetic tape via an interface and calculator for later calculation, using the coordinates of the two standards (high and low) and the power model (y = a. xb) which in a double logarithmic coordinate system gives a straight line relationship (In y = b In a). The calculator (Hewlett Packard 9815 A) corrects automatically for base-line drift, and on the basis of the stored data calculates the sample values from the peak height differences. Precision: The precision of the method is “Within-day” C.V.% = 3.6% and “between-day” C.V.% = 9.9%. Accuracy: The accuracy is determined by means of comparison studies and
325
Fig. 1. The correlation of pregnancy zone protein methods, rocket electrophoresis and nephelometry.
(PZP)
concentration
measured
by two immunochemical
recovery studies. The recovery studies include 20 different patient samples analysed before and after adding of 71 mg/l of PZP from a reference serum. On an average the recovery was 86 mg/l (121%) with an observed range of 62-135 mg/l(72%-157%). Electra immunoassay Rocket electrophoresis is carried out by the method of Laurel1 agarose gel in sodium barbital pH 8.6, 0.017 M with an antiserum tion of 2%. The standard (serum pool 800 mg/l) is used undiluted dard) and diluted 1 + 9 (low standard). The serum samples are treated and undiluted 5 ~1 of each. The electrophoresis is run for Volt. The precipitates are stained with Ponceau S.
[17] in 1% concentra(high stanapplied un16 h at 25
Correlation studies As demonstrated in Fig. 1 there is a positive (b = 0.81) and significant 0.92, p < 0.001) correlation between the AIP method and the EIA method.
(r =
Material The controls (C) consisted of 111 healthy blood donors, 56 women aged 19-60 years and 55 men aged 19-64 years. The patient material included 28 women aged 22-87 years and 32 men aged 35-81 years, suffering from various malignancies. The women mainly had cancer of the breast, lungs or the female genital system. Among the men’s cancer of the lungs and the gastrointestinal tract were predominant. The diagnoses were confirmed by biopsy, during operation or at autopsy. The malignancies were graduated into three stages. Stage I (N,M,): Patients with localized tumours but without regional lymph node metastases or distant metastases. Stage II (NiM,-,): Patients with histologically verified lymph node metastases but without distant metastases. Stage III (NiMi): Patients with distant metastases. Stage I included 7 women
326
and 7 men, stage II 7 women and 10 men and stage III 14 women and 15 men. None of the women were pregnant and nor were any of the patients being treated with oestrogens. For statistical evaluations the non-parametric Mann-Whitney test was employed. We consider p values CO.01 as significant, 0.01 < p < 0.05 as possibly significant, and p > 0.05 as insignificant. Results The PZP values in sera from healthy subjects and in the three different stages of various malignancies are shown in Table I and in Fig. 2. Women: The interval of the PZP values in the three stages of malignancy is almost identical with the 0.95 reference interval of the controls. In stage I all observations are within the 0.95 reference interval, while 2 out of 7 in stage II and 3 out of 14 in stage III are above the 0.95 reference interval. On the other hand 8 out of 56 healthy women have PZP values below the interval of women with malignant disease.
RZ .P. mg/l
.
Female
. .lOO
. :
.
.
;
. __-.---
.
.
68
1 C f?Z.P.
I
__&__ . .
N*M,
N M.
NM,
Male
mg/f 100
Fig. 2. Serum PZP values of healthy male and female controls (C) related to patients with various malig. nancies. Localized tumour is indicated by NoMo. tumour with regional lymph node metastases by NlMo and disseminated tumour by N1 Ml. The median values are indicated by horizontal dot-and-dash lines.
321
TABLE I THE AVERAGE CONCENTRATION (MEDIAN) AND INTERVAL DONORS (CONTROLS) AND IN PATIENTS WITH DIFFERENT
OF PZP IN HEALTHY STAGES OF VARIOUS
BLOOD MALIG-
NANCIES Reference interval. 0.95. Females n
Controls (C) Stage I (NoMo). localized tumours Stage II (N, MO), regional lymph node metastases Stage III (N1 Ml ). distant metastases
Males Median
Interval
(mgP)
cmgn,
56 7 7
38 34
a4
2-91 19-87 21-132
14
86
13-99
n
Median
Interval
(mg/I)
(mgP)
55 7 10
2 11 12
O-20 O-17 o-43
15
47
o-93
Table II shows the significant differences of the PZP values between the different stages. The PZP values in patients with stage II and stage III tumours are increased significantly compared with the controls, while possible significance is demonstrated versus stage I tumours. Men: All PZP values in stage I are within the 0.95 reference interval. In stage II 3 out of 10 and in stage III 11 out of 15 are above the 0.95 reference interval. As demonstrated in Table II the PZP values in stage III tumours are significantly increased compared with controls, stage I and stage II tumours as well. Possible significance is demonstrated between the controls and stage II tumours. No correlation between PZP values and age was demonstrated either in the female or in the male material. Discussion In accordance with other investigations [2,4,14] we find higher PZP values in serum from non-pregnant, non-oestrogen treated women (38 mg/l) compared with healthy men (2 mg/l). The PZP values in cancer patients reported in previous publications have TABLE II THE SIGNIFICANT DIFFERENCES OF THE PZP VALUES CIES COMPARED INDIVIDUALLY AND WITH CONTROLS No significant difference is indicated by -,
I vs. controls II vs. stage I II vs. controls III vs. stage II III vs. stage I III vs. controls
Females
Males
-
-
(+I + (+) +
STAGES
OF MALIGNAN-
possibly significant by (+) and significant by +.
Statistical significance
Stage Stage Stage Stage Stage Stage
IN VARIOUS
(+) + + +
_
328
been contradictory. Some authors have demonstrated elevated values in both sexes, specially in cancers with distant metastases [5,13,15]. Berne [14] reported, that females frequently develop detectable elevation of PZP values in cancer, but males do not. Finally, some investigators have been unable to demonstrate elevated PZP values in cancer patients at all. Our results seem to confirm a relationship between increase in the PZP values and spreading of the malignant disease in both sexes. The diagnostic value of a single PZP determination concerning malignancy contra no malignancy and in differentiation between various tumour stages appears limited. We found the highest diagnostic value in men with distant metastases. PZP values higher than 20 mg/l indicate distant metastases. On the other hand serum PZP values below 20 mg/l in men do not exclude distant metastases. In women PZP values higher than 100 mg/l indicate a malignant disease with either regional lymph node metastases or distant metastases, while PZP values below 10 mg/l seem to exclude tumours. In conclusion our results demonstrate: (1) In cancer detection the diagnostic value of a single PZP determination is limited. (2) A disseminated malignant disease must be suspected in males with a PZP value higher than 20 mg/l. (3) Longitudinal studies of the PZP values may be of considerable value in monitoring cancer patients or as a residue control. Serial PZP determinations in patients with different malignancies in various tumour stages are under investigation. References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Smithies, 0. (1959) Adv. Protein Chem. 14,65 Bohn, H. (1974) Arch. Gyniikol. 217.219-231 Stimson, W.H. and Sinclair. S.M. (1974) FEBS L&t. 47,190 Von Schoultz, B., Damher. M.-G. and St&brand. T. (1976) Protides of Biological Fluids (Peeters. H., ed.). Vol. 24, pp. 189-192, Pergamon Press, Oxford Home, C.H.W., McLay, A.L;C., Tavadia. H.B.. Carmichael, I., MaIlinson, A.C., Young Laiwah, A.A.C., Thomas, M.A. and MacSween, R.N.M. (1973) CIii. Exp. Immunol. 13,603--611 Afonso, J.F. and de Alvarez. R.R. (1963) Am. J. Obstet. Gynecol. 86,815 Cooper, D.W. (1963) Nature 200.892 Beckman, L.. Von Schoultz. B. and Stigbrand, T. (1973) Acta Obstet. Gynecol. Stand. 51. 157-160 Beckman, L.. Von Schoultz. B. and Stigbrand. T. (1971) Acta Obstet. Gynecol. Stand. 50, 369371 Hagen. C. and Fr#land. A. (1972) Lancet i. 1185 (Letter to the editor) Von Schoultz, B., Stigbrand. T. and Tiimvik. A. (1973) FEBS Lett. 38.23 Than, G.N., Csaba, F., Szab6, D.G.. Karg, N.J. and NovPk, P.F. (1975) Arch. Gynlkol. 218, 125130 Stimson. W.H. (1975) J. CIin. Pathol. 28.868-871 Berne. B.H. (1976) Protides of Biological Fluids (Peeters. H.. ed.). Vol. 24, PP. 165-179. Pergamon Press. Oxford Anderson, J.M.. Stimson, W.H. and Kelly. F. (1976) Br. J. Surg. 63.819-822 HeUsing. K. and EngstrBm, H. (1977) Stand. J. Clin. Lab. Invest. 37, 529-536 LaureII, C.-B. (1972) Stand. J. CIii. Lab. Invest. 29, Suppl. 124,21-37
