Clin. exp. Immunol. (1979) 36, 317-325.
Elevated serum levels of a biliary glycoprotein (BGP I) in patients with liver or biliary tract disease T. SVENBERG,* B. WAHRENt & S. HAMMARSTROM+ *Department of Clinical Chemistry, Danderyd Hospital, Danderyd, tDepartment of Virology, National Bacteriological Laboratory, Stockholm, Department of Immunology, University ofStockholm, S-106 91 Stockholm, Sweden
(Accepted for publication 2 October 1978)
SUMMARY
Human hepatic bile contains a glycoprotein (biliary glycoprotein I, BGP I) which cross-reacts with the carcinoembryonic antigen (CEA). A radioimmunoassay for BGP I was developed. The interference of CEA or 'non-specific cross-reacting antigen' (NCA) in the assay was small. The serum levels of BGP I were determined in healthy subjects, in patients with hepato-biliary diseases and in patients with various infectious or inflammatory disorders. Healthy individuals, including pregnant women, had a serum BGP I concentration of about 0-5-1 mg/l. Diseases of the liver or biliary tract (e.g. hepatitis A or B, cytomegalovirus hepatitis, obstructive jaundice or primary biliary cirrhosis) were associated with elevated serum levels of BGP I, as opposed to infectious diseases not affecting the liver mostly showing values within the normal range. Raised levels of serum BGP I activity may reflect biliary obstruction as a result of interference with normal BGP I secretion to the bile.
INTRODUCTION Human bile contains a glycoprotein called biliary glycoprotein I, (BGP I) (Svenberg, 1976). The concentration of BGP I in hepatic bile from non-icteric patients was 5-10 mg/l. In contrast, the biliary concentration of BGP I in patients with jaundice due to pancreatic cancer was low or undetectable (< 1 mg/l) (Svenberg, Hammarstr6m & Hedin, 1978). BGP I was purified from hepatic bile and characterized chemically and immunologically. It is closely related to carcinoembryonic antigen (CEA) (Gold & Freedman, 1965) and to 'non-specific cross-reacting antigen' (NCA) (von Kleist, Chavanel & Burtin, 1972; Hammarstrom et al., 1978; Svenberg et al., 1978). The molecular weight of BGP I is approximately 90,000 Daltons. Immunofluorescence studies of normal human gastrointestinal tissues showed that BGP I is present only in the biliary tract (Svenberg, Hammarstrdm & Zeromski, 1979). No BGP I fluorescence was seen in hepatocytes or other cells lining the biliary tract. Whether BGP I is produced by the hepatocytes or is produced elsewhere and secreted by the liver to bile is not yet known. In either case, diseases of the liver or biliary tract would be expected to cause impairment of BGP I secretion to the bile. This, in turn, could result in the accumulation of BGP I in serum, which would thus have diagnostic implications. In the present study a method for the serum determination of BGP I is described and evaluated by studying healthy subjects, patients with different hepato-biliary diseases and patients with diseases not affecting the liver. The data indicate that impairment of liver function results in elevated serum levels of BGP I, probably as a result of biliary obstruction. *Present address and correspondence: Department of Surgery, Karolinska Hospital, S-104 01 Stockholm, Sweden. 0099-9104/79/0050-0317 $02.00 ©) 1979 Blackwell Scientific Publications 317
318
T. Svenberg, B. Walren & S. Harnrarstriim MATERIALS AND METHODS
Patients. Sera from the following patient groups were studied: hepatitis A (n 18), hepatitis B (n 20), cytomegalovirus hepatitis (CMV) (n 18), primary biliary cirrhosis (n = 5), obstructive jaundice (n 15), acute infectious or inflammatory disease (n = 35) and healthy persons (n = 30). The diagnoses of hepatitis A and B were obtained from epidemiological data and from the results of HBsAg determinations (Magnius, 1975). Most sera were kindly provided by Dr L. 0. Magnius, National Bacteriological Laboratory, Stockholm. CMV infections were determined by sero-conxversion in complement fixation and clinical data. All sera were from the National Bacteriological Laboratories. CMV infection gives a more or less pronounced liver affection in 70-90% of cases (Gadler & Wahren, 1978). BGP I assays were made on acute sera from patients with infectious liver disease. Primary biliary cirrhosis was diagnosed by microscopic examination of liver biopsies and according to the criteria of Leevy, Popper & Sherlock (1976). These sera were kindly provided by Dr C. Johansson, Department of Medicine, Karolinska Hospital, Stockholm. Seven of the fifteen patients with obstructive jaundice had gallstone(s) in the common bile duct, the remaining eight had pancreatic cancer. The diagnoses were obtained at laparaotmy in all cases. Sera were obtained before, or within 10 days after operation, i.e. common duct exploration and removal of gallstones or application of a temporary percutaneous bile drain in cases of pancreatic cancer. All patients with acute infectious or inflammatory disease had more or less pronounced serum protein abnormalities (acute phase reaction) and normal lix er tests (see below). The age distribution in this group was 6-87 years, eighteen patients were 50 years or older. The diagnoses were pneumonia (ii 10), pyelonephritis (ai 3), tonsillitis (nl 4), meningitis (?i 4), 3), mumps (n 2), erysipelas (n 3), toxoplaspolymyalgia rheumatica (nl = 2), septic arthritis (n = 2), septicaemia (n mosis (n 1) and thyreoiditis (n = 1). Sera from healthy individuals were from young adults of both sexes (twenty male blood donors and ten women in early pregnancy). Sera from four patients with skeletal metastases of mammary or prostatic cancer and from six women in late pregnancy, and thoracic duct lymph from four patients with myasthenia gravis were also studied. Sera from the former two groups contained increased levels of bone and placental alkaline phosphatase (ALP), respectively. Thoracic duct lymph is known to be rich in intestinal ALP (Keilding, 1966). The lymph specimens were obtained from Dr A.K. Lefvert, Karolinska Hospital, Stockholm. Biliary glycoprotein I. BGP I was purified from normal human hepatic bile by a five step procedure described in detail elsewhere (Svenberg et al., 1978). Briefly, dialysed and centrifuged hepatic bile was precipitated with perchloric acid (PCA) (1-0 M) and the soluble material subjected to ion-exchange chromatography on DEAE-Sephadex A50, concanavalin ASepharose 4B affinity chromatography, followed by two successive gel filtrations (Sephadex G200 and Sepharyl S200 in guanidine-HCl). Purified BGP I labelled with I251 phosphate buffered saline (PBS), or Sepharose 4B in 6 0 gave a narrow band on sodium-dodecylsulphate polyacrylamid gel electrophoresis and was precipitated to 80-85% by BGP I or CEA antisera (Svenberg et al., 1978). BGP I antibodies. Rabbits were injected intramuscularly with purified BGP I in Freund's complete adjus ant (50- 100jug of BGP I per injection) exery third week. Antiserum collected after three injections was used. It was absorbed twice with human AB erythrocytes and purified on a cross-reacting CEA immunoadsorbent (CEA-IS) as described previously (Svenberg et al., 1978). Briefly, 30 ml of antiserum was sequentially passed oxer a CEA-IS. At each step the bound antibodies were eluted with 6-0 NI guandine-HCl and the passed fraction added to CEA-IS. Different amounts of the same CEA-IS were used in the consecutive steps (twice with 5 0 mg, followed by twice with 80 mg). The BGP I antibodies eluted from the last CEA immunoadsorbent step were used in the radioimmunoassay. The rationale for selecting this antibody fraction was (1) to ensure that our BGP I antibodies were CEA cross-reactive (BGP I is defined by its immunological cross-reactixity with CEA) and (2) to obtain BGP I antibodies with a high degree of selectix ity for BGP I as compared to CEA and NCA. When tested in immunodiffusion, this antibody fraction precipitated with BGP I but not with CEA or NCA (Svenberg et al., 1978). In the BGP I radioimmunoassay (see below) 50% inhibition was obtained with 20 ng BGPI; 200-10,000 ng NCA and 1000 ng CEA (Fig. 1). CE.4 and NCAI. CEA and NCA were purified from liver metastases of colo-rectal cancer. NCA was also purified from normal adult spleen (spleen NCA). The purification procedures and properties of these materials ha-e been described previously (Hammarstrom, Engxall & Sundblad, 1976; Hammarstrom et al., 1978). Anti-BGP I immunoadsorbent. The IgG fraction, obtained by protein A-Sepharose 4B (Pharmacia, Uppsala, Sweden) fractionation of a high titred rabbit BGP I antiserum, was coupled to CNBr-activated Sepharose 4B (Pharmacia, Uppsala, Sweden). 200 mg of IgG containing about 10 mg of specific antibody was conjugated to 35 ml of gel. 2-0-5 0 ml aliquots of patient sera were fractionated on columns containing 4 0 ml of gel. The columns were washed with PBS andoluted with 6 0 guanidine-HCl. I labelled with Na ["25I] (>14 mCi jig, carrier-free, RadioBGP I radioimminioassay (BGP IRIL). Purified BGP was jig of BGP I was labelled chemical Centre, Amersham) using the chloramine T procedure (Hunter & Greenwood, 1962). 50 with 2 0 mCi of Na [I251]. Labelled BGP I had a specific activity of .20 jiCi/lig. BGP I determinations were performed with a double antibody assay of the same type as used for CEA and NCA (Wahren, Edsmyr & Zimmerman, 1975; Gadler al., 1978). It was performed as follows: 0 2 ml of standard BGP I solution (15-100 ngBG I P ) or the sample dilution to be assayed was mixed with 0 ml of BGP I antibodies (see aboxe) at a final concentration of 0 1 tg/ml in PBS containing 1% boxine serum albumin (BSA), and incubated for 6 hr at 37cC. 0 2 ml PBS and 0 1 ml of 25I]-BGP I (3-0-5104 ct/min) was then added and the mixture incubatedovernight at room temperature. 1 0 ml of a =
=
=
=
=
M
Ns
et
1
Serum levels of BGP I
319
suspension ofcellulose-bound sheep anti-rabbit IgG (Organon, Oss, Holland), diluted 1: 50, w as added followed by incubation at 37 C for 2 hr with intermittent agitation. The mixture was then centrifuged at 1000 g for 5 min and the pellet washed once with PBS-BSA. The pellet was finally counted in a Packard 2-counter. At the dilution of BGP I antibodies used in the assay, about 40%o of total radioacti-e tracer was bound (Fig. 1). Both the inter- and intra-assay coefficients of variation, defined as 1 standard deviation in percent of mean, were 14%. CEA- and NCA-RIA. These were performed as described by Wahren et al. (1975) and Gadler et al. (1978). In both assays, non-PCA extracted serum specimens were tested. Liver tests. Total bilirubin concentration in serum (S-Bil, reference range: 4-21 ,mol/l) was determined in a Technicon autoanalyser according to Gambino & Schreiber (1964). Serum activities of aspartate-aminotransferase (S-ASAT, S-ALAT, reference values < 0 70 ,ukat/l) and alkaline phosphatase (S-ALP, reference range: 0 8-42 ,ukat/l) were assayed in a Vitatron AKES II automatic analyser according to the Committee on Enzymes of the Scandinavian Society for Clinical Chemistry and Clinical Physiology (1974). Gammaglutamyltranspeptidase activity in serum (S-GT, reference ranges: males: 0 17-1 00, females: 0 17-034 ,ukat/l) wxas determined according to the above mentioned Committee (1976) in a LKB 8600 automatic recording photometer. Alkaline phosphatase isoenzymes. The separation of isoenzymes was performed according to HaIgerstrand & Skude (1976) on specimens with elevated ALP of non-hepatic origin. Statistical analysis. To test the significance of the correlations between serum activity of BGP I and different livter tests, a -2 where r = correlation coefficient, n = number of data t values were calculated according to the expression t LI -r2 , pairs and n- 2 = the degrees of freedom.
RESULTS Sensitivity and specificity of the BGP I-RIA Fig. 1 shows the inhibition curves obtained with BGP I, CEA and NCA. Individual BGP I preparations gave parallel inhibition curves (sp. act. 70-125% of the preparation shown). The least detectable quantity of BGP I was about 5 ng. Individual CEA preparations were inhibitory at 50-100 times higher concentrations than BGP I (one preparation shown in Fig. 1). NCA inhibited at a 10-> 500 times higher concentration than BGP I. Two of the NCA preparations, tumour NCA (45) and tumour NCA (41),
0001
001
0-
10
(pg) FIG. 1. Specificity of BGP I radioimmunoassay. (0) Standard BGP I; ( ) CEA;(() spleen NCAhjigh; (-) spleen NCA70W; (At) tumour NCAh jgh (41); (A) tumour NCAIOW (41) (*) tumour NCA (45). For details on NCA preparations see Hammarstrom et al. (1978). Inhibitor added
high and low molecular weight material (Hammarstrom et al., 1978) were extremely poor inhibitors. In contrast, the high and low molecular weight materials of NCA from pooled adults spleens were relatively good inhibitors. Table 1 shows the levels of BGP I, NCA and CEA in individual sera from patients with acute hepatitis (n = 20), and biliary obstruction (n = 4). The mean values for BGP I, NCA and CEA in healthy
320
T. Svenberg, B. Wahren & S. Hammarstrim TABLE 1. BGP I NCA and CEA levels in sera from patients with acute hepatitis and obstructive jaundice BGP I (mg/i)
NCA (mg/I)
Hepatitis A
1-47 2*00 3-91 1-47 2-55 2*73 1*87 4-48 1-50 3-58
Hepatitis B
2-67 4-44 2-37 2-65
2-30
0-14 0-18 0-21 0*18 0-22 0-28 0-19 0-22 0-19 0-23 0-23 0-16 0-19 0-13 0-16 0 30 0-22 0-32
2*49
0*15
Diagnosis
3*04 2-00
3*06 4-11 2-55
0-22 Obstructive jaundice 0-06 0 09 5*75 2-65 0-13 2-94 0 09 Healthy individuals 0-67+0.27* 0-15+0-12 (n = 30) (n = 20)
CEA
(ug/I) 28 24 29 27 27 37 23 27 28 29 28 34 28 26 33 31 27 33 31 36 33 58 101 42 24±7 (n = 19)
* Mean+ 2 s.d.
individuals are also given. Sera were diluted 1:40 before testing in the BGP I assay, but run undiluted in the CEA- and NCA-RIAs. As can be seen, the serum concentration of NCA was between 10-50 times lower, and the concentration of CEA about 100 times lower, than that of BGP I. BGP I was three- to six-fold elevated in all samples, while only three out of twenty-four patients had slightly elevated NCA. Eight out of twenty-four had slightly raised CEA values. Thus, while BGP I was strongly elevated in most samples, only marginal elevations of NCA and CEA were found, except for CEA in one case of pancreatic cancer (101 pg/l). Serum BGP I (S-BGP I) levels in patients with or without hepato-biliary disease Fig. 2 shows the result of serum BGP I determinations in patients with liver or biliary tract disease, in healthy individuals and in patients with acute infections or inflammations without evidence of liver disease. Healthy donors had S-BGP I values between 0-42 and 0-96 mg/l (mean value+ 2 s.d. = 0-67+ 0-27 mg/l, n = 30). All patients with acute hepatitis A or B had elevated S-BGP I levels, e.g. values > 1 0 mg/l. The range was 1-5-4-9 mg/l, n = 38. In the CMV group, fifteen out of eighteen patients had elevated S-BGP I. Two of the three CMV patients with normal S-BGP I values had normal liver tests. Among the CMV patients with elevated S-BGP I values, four had normal liver tests. All fifteen patients with obstructive jaundice had elevated S-BGP I levels (range 1 9-8 5 mg/l). Patients with common bile duct stones or pancreatic cancer did not differ significantly in their S-BGP I values. All five patients with primary biliary cirrhosis had highly elevated S-BGP I values (range 3-0-6-5 mg/l). In contrast, S-BGP I
Serum levels of BGP I
321
8 7 *
4 -*
Q_
co
VI)
~
'
3
~
0
0~~~~~~
o~~~
>, M
amg.
0
E aE aIliverenzes I
I
~~~~~~ ._ _
._
T
levels of
~
i
D
(_3
I
a.
c
FIG. 2. Serum BGP I (S-BGP I) levels in healthy individuals, patients with diseases of the liver or biliary tract and
in
patients with acute
infectious
or
inflammatory
disease without evidence of liver involvement.
levels were normal or only slightly increased in a group of thirty-five patients with acute infections or inflammations without evidence of liver disease (range
0A-1F5
mg/l, twelve out of thirty-five patients
> 10 mg/l).
Correlation between serum levels of BGP I, different liver enzymes and bilirubin 5-Bil, S-ALP, 5-ASAT and S-ALAT were determined in fifty-six patients with infectious liver disease and in twenty patients with biliary obstruction (obstructive jaundice and primary biliary cirrhosis). The
TABLE levels
2.
Correlation
of BGP
I,
liver
coefficients enzymes
between and
serum
bilirubin
in
patients with hepatitis (A, B and CMV, n = 56) and
biliary obstruction (obstructive jaundice, n = 15 and primary biliary cirrhosis, n = 5)
S-BGP I/S-BIL S-BGP I/S-ALP S-BGP I/S-ASAT S-BGP I/S-ALAT *
Hepatitis
Biliary obstruction
0.43 * 0*68t
-0 124 0-80t
0-38* 0 46t
0-381 0-311
P< 0-005.
t P< 0-0005. Not significant.
levels were compared with S-BGP I. Table 2 shows the calculated correlation coefficients and statistical significances for the comparison between S-BGP I on the one hand, and liver enzymes or bilirubin on the other. The correlation between S-BGP I and S-ALP is statistically highly significant in both patients groups (P< 0 0005). In patients with hepatitis there was a highly significant correlation between S-BGP I and S-ALAT (P
Elevated serum levels of a biliary glycoprotein (BGP I) in patients with liver or biliary tract disease T. SVENBERG,* B. WAHRENt & S. HAMMARSTROM+ *Department of Clinical Chemistry, Danderyd Hospital, Danderyd, tDepartment of Virology, National Bacteriological Laboratory, Stockholm, Department of Immunology, University ofStockholm, S-106 91 Stockholm, Sweden
(Accepted for publication 2 October 1978)
SUMMARY
Human hepatic bile contains a glycoprotein (biliary glycoprotein I, BGP I) which cross-reacts with the carcinoembryonic antigen (CEA). A radioimmunoassay for BGP I was developed. The interference of CEA or 'non-specific cross-reacting antigen' (NCA) in the assay was small. The serum levels of BGP I were determined in healthy subjects, in patients with hepato-biliary diseases and in patients with various infectious or inflammatory disorders. Healthy individuals, including pregnant women, had a serum BGP I concentration of about 0-5-1 mg/l. Diseases of the liver or biliary tract (e.g. hepatitis A or B, cytomegalovirus hepatitis, obstructive jaundice or primary biliary cirrhosis) were associated with elevated serum levels of BGP I, as opposed to infectious diseases not affecting the liver mostly showing values within the normal range. Raised levels of serum BGP I activity may reflect biliary obstruction as a result of interference with normal BGP I secretion to the bile.
INTRODUCTION Human bile contains a glycoprotein called biliary glycoprotein I, (BGP I) (Svenberg, 1976). The concentration of BGP I in hepatic bile from non-icteric patients was 5-10 mg/l. In contrast, the biliary concentration of BGP I in patients with jaundice due to pancreatic cancer was low or undetectable (< 1 mg/l) (Svenberg, Hammarstr6m & Hedin, 1978). BGP I was purified from hepatic bile and characterized chemically and immunologically. It is closely related to carcinoembryonic antigen (CEA) (Gold & Freedman, 1965) and to 'non-specific cross-reacting antigen' (NCA) (von Kleist, Chavanel & Burtin, 1972; Hammarstrom et al., 1978; Svenberg et al., 1978). The molecular weight of BGP I is approximately 90,000 Daltons. Immunofluorescence studies of normal human gastrointestinal tissues showed that BGP I is present only in the biliary tract (Svenberg, Hammarstrdm & Zeromski, 1979). No BGP I fluorescence was seen in hepatocytes or other cells lining the biliary tract. Whether BGP I is produced by the hepatocytes or is produced elsewhere and secreted by the liver to bile is not yet known. In either case, diseases of the liver or biliary tract would be expected to cause impairment of BGP I secretion to the bile. This, in turn, could result in the accumulation of BGP I in serum, which would thus have diagnostic implications. In the present study a method for the serum determination of BGP I is described and evaluated by studying healthy subjects, patients with different hepato-biliary diseases and patients with diseases not affecting the liver. The data indicate that impairment of liver function results in elevated serum levels of BGP I, probably as a result of biliary obstruction. *Present address and correspondence: Department of Surgery, Karolinska Hospital, S-104 01 Stockholm, Sweden. 0099-9104/79/0050-0317 $02.00 ©) 1979 Blackwell Scientific Publications 317
318
T. Svenberg, B. Walren & S. Harnrarstriim MATERIALS AND METHODS
Patients. Sera from the following patient groups were studied: hepatitis A (n 18), hepatitis B (n 20), cytomegalovirus hepatitis (CMV) (n 18), primary biliary cirrhosis (n = 5), obstructive jaundice (n 15), acute infectious or inflammatory disease (n = 35) and healthy persons (n = 30). The diagnoses of hepatitis A and B were obtained from epidemiological data and from the results of HBsAg determinations (Magnius, 1975). Most sera were kindly provided by Dr L. 0. Magnius, National Bacteriological Laboratory, Stockholm. CMV infections were determined by sero-conxversion in complement fixation and clinical data. All sera were from the National Bacteriological Laboratories. CMV infection gives a more or less pronounced liver affection in 70-90% of cases (Gadler & Wahren, 1978). BGP I assays were made on acute sera from patients with infectious liver disease. Primary biliary cirrhosis was diagnosed by microscopic examination of liver biopsies and according to the criteria of Leevy, Popper & Sherlock (1976). These sera were kindly provided by Dr C. Johansson, Department of Medicine, Karolinska Hospital, Stockholm. Seven of the fifteen patients with obstructive jaundice had gallstone(s) in the common bile duct, the remaining eight had pancreatic cancer. The diagnoses were obtained at laparaotmy in all cases. Sera were obtained before, or within 10 days after operation, i.e. common duct exploration and removal of gallstones or application of a temporary percutaneous bile drain in cases of pancreatic cancer. All patients with acute infectious or inflammatory disease had more or less pronounced serum protein abnormalities (acute phase reaction) and normal lix er tests (see below). The age distribution in this group was 6-87 years, eighteen patients were 50 years or older. The diagnoses were pneumonia (ii 10), pyelonephritis (ai 3), tonsillitis (nl 4), meningitis (?i 4), 3), mumps (n 2), erysipelas (n 3), toxoplaspolymyalgia rheumatica (nl = 2), septic arthritis (n = 2), septicaemia (n mosis (n 1) and thyreoiditis (n = 1). Sera from healthy individuals were from young adults of both sexes (twenty male blood donors and ten women in early pregnancy). Sera from four patients with skeletal metastases of mammary or prostatic cancer and from six women in late pregnancy, and thoracic duct lymph from four patients with myasthenia gravis were also studied. Sera from the former two groups contained increased levels of bone and placental alkaline phosphatase (ALP), respectively. Thoracic duct lymph is known to be rich in intestinal ALP (Keilding, 1966). The lymph specimens were obtained from Dr A.K. Lefvert, Karolinska Hospital, Stockholm. Biliary glycoprotein I. BGP I was purified from normal human hepatic bile by a five step procedure described in detail elsewhere (Svenberg et al., 1978). Briefly, dialysed and centrifuged hepatic bile was precipitated with perchloric acid (PCA) (1-0 M) and the soluble material subjected to ion-exchange chromatography on DEAE-Sephadex A50, concanavalin ASepharose 4B affinity chromatography, followed by two successive gel filtrations (Sephadex G200 and Sepharyl S200 in guanidine-HCl). Purified BGP I labelled with I251 phosphate buffered saline (PBS), or Sepharose 4B in 6 0 gave a narrow band on sodium-dodecylsulphate polyacrylamid gel electrophoresis and was precipitated to 80-85% by BGP I or CEA antisera (Svenberg et al., 1978). BGP I antibodies. Rabbits were injected intramuscularly with purified BGP I in Freund's complete adjus ant (50- 100jug of BGP I per injection) exery third week. Antiserum collected after three injections was used. It was absorbed twice with human AB erythrocytes and purified on a cross-reacting CEA immunoadsorbent (CEA-IS) as described previously (Svenberg et al., 1978). Briefly, 30 ml of antiserum was sequentially passed oxer a CEA-IS. At each step the bound antibodies were eluted with 6-0 NI guandine-HCl and the passed fraction added to CEA-IS. Different amounts of the same CEA-IS were used in the consecutive steps (twice with 5 0 mg, followed by twice with 80 mg). The BGP I antibodies eluted from the last CEA immunoadsorbent step were used in the radioimmunoassay. The rationale for selecting this antibody fraction was (1) to ensure that our BGP I antibodies were CEA cross-reactive (BGP I is defined by its immunological cross-reactixity with CEA) and (2) to obtain BGP I antibodies with a high degree of selectix ity for BGP I as compared to CEA and NCA. When tested in immunodiffusion, this antibody fraction precipitated with BGP I but not with CEA or NCA (Svenberg et al., 1978). In the BGP I radioimmunoassay (see below) 50% inhibition was obtained with 20 ng BGPI; 200-10,000 ng NCA and 1000 ng CEA (Fig. 1). CE.4 and NCAI. CEA and NCA were purified from liver metastases of colo-rectal cancer. NCA was also purified from normal adult spleen (spleen NCA). The purification procedures and properties of these materials ha-e been described previously (Hammarstrom, Engxall & Sundblad, 1976; Hammarstrom et al., 1978). Anti-BGP I immunoadsorbent. The IgG fraction, obtained by protein A-Sepharose 4B (Pharmacia, Uppsala, Sweden) fractionation of a high titred rabbit BGP I antiserum, was coupled to CNBr-activated Sepharose 4B (Pharmacia, Uppsala, Sweden). 200 mg of IgG containing about 10 mg of specific antibody was conjugated to 35 ml of gel. 2-0-5 0 ml aliquots of patient sera were fractionated on columns containing 4 0 ml of gel. The columns were washed with PBS andoluted with 6 0 guanidine-HCl. I labelled with Na ["25I] (>14 mCi jig, carrier-free, RadioBGP I radioimminioassay (BGP IRIL). Purified BGP was jig of BGP I was labelled chemical Centre, Amersham) using the chloramine T procedure (Hunter & Greenwood, 1962). 50 with 2 0 mCi of Na [I251]. Labelled BGP I had a specific activity of .20 jiCi/lig. BGP I determinations were performed with a double antibody assay of the same type as used for CEA and NCA (Wahren, Edsmyr & Zimmerman, 1975; Gadler al., 1978). It was performed as follows: 0 2 ml of standard BGP I solution (15-100 ngBG I P ) or the sample dilution to be assayed was mixed with 0 ml of BGP I antibodies (see aboxe) at a final concentration of 0 1 tg/ml in PBS containing 1% boxine serum albumin (BSA), and incubated for 6 hr at 37cC. 0 2 ml PBS and 0 1 ml of 25I]-BGP I (3-0-5104 ct/min) was then added and the mixture incubatedovernight at room temperature. 1 0 ml of a =
=
=
=
=
M
Ns
et
1
Serum levels of BGP I
319
suspension ofcellulose-bound sheep anti-rabbit IgG (Organon, Oss, Holland), diluted 1: 50, w as added followed by incubation at 37 C for 2 hr with intermittent agitation. The mixture was then centrifuged at 1000 g for 5 min and the pellet washed once with PBS-BSA. The pellet was finally counted in a Packard 2-counter. At the dilution of BGP I antibodies used in the assay, about 40%o of total radioacti-e tracer was bound (Fig. 1). Both the inter- and intra-assay coefficients of variation, defined as 1 standard deviation in percent of mean, were 14%. CEA- and NCA-RIA. These were performed as described by Wahren et al. (1975) and Gadler et al. (1978). In both assays, non-PCA extracted serum specimens were tested. Liver tests. Total bilirubin concentration in serum (S-Bil, reference range: 4-21 ,mol/l) was determined in a Technicon autoanalyser according to Gambino & Schreiber (1964). Serum activities of aspartate-aminotransferase (S-ASAT, S-ALAT, reference values < 0 70 ,ukat/l) and alkaline phosphatase (S-ALP, reference range: 0 8-42 ,ukat/l) were assayed in a Vitatron AKES II automatic analyser according to the Committee on Enzymes of the Scandinavian Society for Clinical Chemistry and Clinical Physiology (1974). Gammaglutamyltranspeptidase activity in serum (S-GT, reference ranges: males: 0 17-1 00, females: 0 17-034 ,ukat/l) wxas determined according to the above mentioned Committee (1976) in a LKB 8600 automatic recording photometer. Alkaline phosphatase isoenzymes. The separation of isoenzymes was performed according to HaIgerstrand & Skude (1976) on specimens with elevated ALP of non-hepatic origin. Statistical analysis. To test the significance of the correlations between serum activity of BGP I and different livter tests, a -2 where r = correlation coefficient, n = number of data t values were calculated according to the expression t LI -r2 , pairs and n- 2 = the degrees of freedom.
RESULTS Sensitivity and specificity of the BGP I-RIA Fig. 1 shows the inhibition curves obtained with BGP I, CEA and NCA. Individual BGP I preparations gave parallel inhibition curves (sp. act. 70-125% of the preparation shown). The least detectable quantity of BGP I was about 5 ng. Individual CEA preparations were inhibitory at 50-100 times higher concentrations than BGP I (one preparation shown in Fig. 1). NCA inhibited at a 10-> 500 times higher concentration than BGP I. Two of the NCA preparations, tumour NCA (45) and tumour NCA (41),
0001
001
0-
10
(pg) FIG. 1. Specificity of BGP I radioimmunoassay. (0) Standard BGP I; ( ) CEA;(() spleen NCAhjigh; (-) spleen NCA70W; (At) tumour NCAh jgh (41); (A) tumour NCAIOW (41) (*) tumour NCA (45). For details on NCA preparations see Hammarstrom et al. (1978). Inhibitor added
high and low molecular weight material (Hammarstrom et al., 1978) were extremely poor inhibitors. In contrast, the high and low molecular weight materials of NCA from pooled adults spleens were relatively good inhibitors. Table 1 shows the levels of BGP I, NCA and CEA in individual sera from patients with acute hepatitis (n = 20), and biliary obstruction (n = 4). The mean values for BGP I, NCA and CEA in healthy
320
T. Svenberg, B. Wahren & S. Hammarstrim TABLE 1. BGP I NCA and CEA levels in sera from patients with acute hepatitis and obstructive jaundice BGP I (mg/i)
NCA (mg/I)
Hepatitis A
1-47 2*00 3-91 1-47 2-55 2*73 1*87 4-48 1-50 3-58
Hepatitis B
2-67 4-44 2-37 2-65
2-30
0-14 0-18 0-21 0*18 0-22 0-28 0-19 0-22 0-19 0-23 0-23 0-16 0-19 0-13 0-16 0 30 0-22 0-32
2*49
0*15
Diagnosis
3*04 2-00
3*06 4-11 2-55
0-22 Obstructive jaundice 0-06 0 09 5*75 2-65 0-13 2-94 0 09 Healthy individuals 0-67+0.27* 0-15+0-12 (n = 30) (n = 20)
CEA
(ug/I) 28 24 29 27 27 37 23 27 28 29 28 34 28 26 33 31 27 33 31 36 33 58 101 42 24±7 (n = 19)
* Mean+ 2 s.d.
individuals are also given. Sera were diluted 1:40 before testing in the BGP I assay, but run undiluted in the CEA- and NCA-RIAs. As can be seen, the serum concentration of NCA was between 10-50 times lower, and the concentration of CEA about 100 times lower, than that of BGP I. BGP I was three- to six-fold elevated in all samples, while only three out of twenty-four patients had slightly elevated NCA. Eight out of twenty-four had slightly raised CEA values. Thus, while BGP I was strongly elevated in most samples, only marginal elevations of NCA and CEA were found, except for CEA in one case of pancreatic cancer (101 pg/l). Serum BGP I (S-BGP I) levels in patients with or without hepato-biliary disease Fig. 2 shows the result of serum BGP I determinations in patients with liver or biliary tract disease, in healthy individuals and in patients with acute infections or inflammations without evidence of liver disease. Healthy donors had S-BGP I values between 0-42 and 0-96 mg/l (mean value+ 2 s.d. = 0-67+ 0-27 mg/l, n = 30). All patients with acute hepatitis A or B had elevated S-BGP I levels, e.g. values > 1 0 mg/l. The range was 1-5-4-9 mg/l, n = 38. In the CMV group, fifteen out of eighteen patients had elevated S-BGP I. Two of the three CMV patients with normal S-BGP I values had normal liver tests. Among the CMV patients with elevated S-BGP I values, four had normal liver tests. All fifteen patients with obstructive jaundice had elevated S-BGP I levels (range 1 9-8 5 mg/l). Patients with common bile duct stones or pancreatic cancer did not differ significantly in their S-BGP I values. All five patients with primary biliary cirrhosis had highly elevated S-BGP I values (range 3-0-6-5 mg/l). In contrast, S-BGP I
Serum levels of BGP I
321
8 7 *
4 -*
Q_
co
VI)
~
'
3
~
0
0~~~~~~
o~~~
>, M
amg.
0
E aE aIliverenzes I
I
~~~~~~ ._ _
._
T
levels of
~
i
D
(_3
I
a.
c
FIG. 2. Serum BGP I (S-BGP I) levels in healthy individuals, patients with diseases of the liver or biliary tract and
in
patients with acute
infectious
or
inflammatory
disease without evidence of liver involvement.
levels were normal or only slightly increased in a group of thirty-five patients with acute infections or inflammations without evidence of liver disease (range
0A-1F5
mg/l, twelve out of thirty-five patients
> 10 mg/l).
Correlation between serum levels of BGP I, different liver enzymes and bilirubin 5-Bil, S-ALP, 5-ASAT and S-ALAT were determined in fifty-six patients with infectious liver disease and in twenty patients with biliary obstruction (obstructive jaundice and primary biliary cirrhosis). The
TABLE levels
2.
Correlation
of BGP
I,
liver
coefficients enzymes
between and
serum
bilirubin
in
patients with hepatitis (A, B and CMV, n = 56) and
biliary obstruction (obstructive jaundice, n = 15 and primary biliary cirrhosis, n = 5)
S-BGP I/S-BIL S-BGP I/S-ALP S-BGP I/S-ASAT S-BGP I/S-ALAT *
Hepatitis
Biliary obstruction
0.43 * 0*68t
-0 124 0-80t
0-38* 0 46t
0-381 0-311
P< 0-005.
t P< 0-0005. Not significant.
levels were compared with S-BGP I. Table 2 shows the calculated correlation coefficients and statistical significances for the comparison between S-BGP I on the one hand, and liver enzymes or bilirubin on the other. The correlation between S-BGP I and S-ALP is statistically highly significant in both patients groups (P< 0 0005). In patients with hepatitis there was a highly significant correlation between S-BGP I and S-ALAT (P
