- Email: [email protected]
Serum bile acid concentrations during pregnancy and their relationship to obstetric cholestasis
GASTROENTEROLOGY 1986;91:825-9
Serum Bile Acid Concentrations During Pregnancy and Their Relationship to Obstetric Cholestasis MICHAEL LUNZER, PHILIP BARNES, KAREN BYTH, and MARY O’HALLORAN Department of Gastroenterology, Royal North Shore Hospital and Department Royal Alexandra Hospital for Children, Sydney, Australia
We hove prospectively studied changes in serum postprandial cholylglycine (CG) concentration during 297 pregnancies. We found an increase in CG concentration from 0.3 mol/L at 15 weeks’ pregnancy to 0.6 PmollL at 40 weeks’ pregnancy. Although this increase was statistically significant (p < 0.005), median concentrations of CG remained well within the normal range (O-l.5 PmollL). However, 10% of the group showed markedly elevated serum CG concentrations at 30 weeks’ pregnancy, and the CC level in this group continued to rise during the third trimester. Pruritus was significantly more common in the group with elevated CG concentrations (48%) than in the group with normal CG levels (20%) (p < 0.005). Serum CG was a much more sensitive predictor of pruritus during pregnancy than other biochemical liver tests. Elevated CG levels were found more commonly in Mediterranean and Asian patients than patients of other ethnic origins [p < 0.025). No statistically signifiwere found between elevated CG cant associations concentrations and maternal age, number of previous pregnancies, pruritus during previous pregnancies, contraceptive-induced cholestasis, and fetal maturity. We conclude [a) that obstetric cholestasis is probably much more common than previously Received August 7, 1985. Accepted April 12, 1986. Address requests for reprints to: Dr. Michael Lunzer, Department of Gastroenterology, Royal North Shore Hospital, St. Leonards, Sydney, N.S.W. 2065, Australia. The authors thank Abbott Diagnostics, North Chicago, Illinois for their generous donation of radioimmunoassay kits for the assessment of serum cholylglycine concentration. The authors thank Elaine Brecknock and Leslie McAdam for collection of data, Dr. Hilary Lunzer for computation of data, and Evelyn Humpherson for manuscript preparation. They also acknowledge the cooperation of the Medical Staff of the Department of Obstetrics at Royal North Shore Hospital and thank them for allowing them to study their patients. 8 1986 by the American Gastroenterological Association 0016-5085/86/$3.50
of Biochemistry,
suspected and (b) that consideration should be given to the measurement of serum bile acids in all pregnant individuals with unexplained pruritus. Obstetric cholestasis is an unusual condition characterized by pruritus and, occasionally, jaundice during the latter part of pregnancy. The symptoms and biochemical abnormalities of cholestasis typically subside soon after birth and tend to recur during subsequent pregnancies (1).An association exists between obstetric cholestasis and oral contraceptive-induced cholestasis, and the condition may be familial (2). Although obstetric cholestasis does not have a deleterious effect on the mother, there are several reports of significant risk to the fetus including prematurity, low birth weight, intrapartum asphyxia, and increased fetal mortality (3-5). Obstetric cholestasis is believed to be a rare condition outside certain Scandinavian and South American countries. Reported prevalence rates for obstetric cholestasis in Chile (12%22%) (6), Bolivia (9%) (7), and Sweden (2%3%) (8) are considerably higher than those found in Australia (O-2%-0.8%) (9,10), France (0.2%) (ll), and Canada (0.1%)(4). No prevalence data are available for the United States, Britain, or most other parts of the world. Mild abnormalities of the routine biochemical liver tests are frequently found in patients with significant obstetric cholestasis. The most marked abnormalities, however, have been reported in the serum concentration of bile acids. Sjovall and Sjovall (12) found 30-fold elevation of serum total bile acids in 6 patients with obstetric cholestasis. The most abnormal elevation of the individual bile acids was that of cholic acid, which was elevated i’O-fold. Similar results were reported by Laatikainen and Ikonen (5). Fulton et al. (l3), in a small series of Abbreviation
used in this paper: CG, cholylglycine.
826
GASTROENTEROLOGY Vol. 91. No. 4
LUNZER ET AL.
26 women, found elevated fasting (but not postprandial) concentrations of cholate after 34 weeks’ gestation. Heikkinen et al. (14) found lo-fold elevations in serum levels of cholic acid in women with obstetric cholestasis, and noted that in 4 of 8 women elevation of serum bile acid levels preceded the appearance of symptoms or other laboratory evidence of intrahepatic cholestasis. There have, however, been no previous reports of a large population on whom serial measurements of serum bile acids have been performed prospectively during pregnancy. In the present study we have chosen to use a commercially available assay for estimating the glycine conjugate of cholic acid. Although the assay measures predominantly the glycine conjugate, the antibody does have cross-reactivity with taurocholic acid (16.5%) and glycochenodeoxycholic acid (14.5%). Some investigators have claimed that the sulfate esters of bile acids represent the major form in hepatobiliary disease; however, the main component of this sulfated group, sulfochenodeoxycholic acid, is not easily assayed in “routine” laboratories. It should also be pointed out that a good correlation exists between the concentration of sulfated bile acids and the concentration of serum glycocholate. The aim of the present study was to define the changes in serum cholylglycine (CG) concentration in pregnant women and to note the relationship of these changes to pruritus and other clinical and biochemical parameters.
Materials and Methods Two hundred ninety-seven pregnant individuals attending the Ante-Natal Clinic at Royal North Shore Hospital were prospectively studied during a 21-mo period from March 1981 to December 1982. All patients attending the clinic during this period were given an opportunity to participate in the study; the 297 participants represented those individuals who agreed to participate after being provided with full details of the requirements of the study. The mean age of participants was 27 yr (range 16-42 yr) and the mean number of previous pregnancies was 1.1. Participants were seen on at least five occasions during pregnancy. The initial visit was at -15 weeks’ gestation, with subsequent visits at -26, 32, 38, and 38 weeks’ gestation. At each visit patients were asked about symptoms of cholestasis (pruritus, dark urine, pale stools, jaundice) and drug ingestion before and during the current pregnancy. Participants were considered to have sustained pruritus if this symptom was present on at least two occasions during late pregnancy (after 28 weeks’ gestation). There were no patients with known hepatobiliary, intestinal, or pancreatic disease in the study. No patients gave a history of alcohol consumption exceeding 40 g daily. No patients were taking drugs known to affect hepatobiliary function. The ethnic origin of participants was also ascertained.
fn i; F
30.
g0 8
II 20.
I3 g L
I --
MEDIAN
= 0.5 JIM
RANGE
=O.l
(N=50)
10.
1
2
1. Two-hour postprandial tions in controls.
4
3
SERUM CHOLYLGLYCINE
Figure
- 1.5 JJM
~JM
serum cholylglycine
concentra-
Participants were asked to have a meal containing a standard amount of fat (18 g) 2 h before their visit, and blood was taken at each visit for assay of serum CG and conventional biochemical liver tests (bilirubin, alkaline phosphatase, -y-glutamyl transferase, aspartate transaminase, alanine transaminase, albumin, globulins). It was not considered justifiable to take multiple samples for serum CG measurement at clinic visits. Blood was also taken for the above tests 48-72 h postpartum. Serum CG was measured by radioimmunoassay (Abbott Diagnostics, North Chicago, Ill.) and “conventional” biochemical liver tests were measured by a S.M.A.C. autoanalyzer. Serum was assayed for hepatitis B surface antigen and immunoglobulin M hepatitis A virus antibody by radioimmunoassay in all patients with elevated serum CG concentrations on sera that showed the highest CG concentration. The computer package S.P.S.S. (Release 8) was used to analyze the data. The 2 and paired t-tests were used, as were linear regression techniques utilizing the F test for linear effect allowing for possible different intercepts for individuals.
Results A control population of 50 age-matched, nonpregnant female volunteers was obtained and data for 2-h postprandial serum CG concentrations in this group are shown in Figure 1. A normal range of o-l.5 pmol/L was found. The changes in serum CG found during pregnancy are shown in Figure 2. Serum CG rose progressively from a median concentration of 0.3 PmollL in early pregnancy to 0.7 pmol/L in late pregnancy, with a fall to 0.6 pmol/L in the immediate postpartum period. Although the rise in CG remained well within the normal range, the increase was statistically significant (p < 0.005). The fall in CG immediately after delivery also reached statistical significance (p < 0.01)
October 1986
SERUM BILE ACIDS DURING PREGNANCY
827
1.5 -
1.2%
1.0 -
Figure
n-t -10
15
2b
25
30
3j
2. Changes in serum cholylglycine during pregnancy (median values].
40
Weeks Pregnant
We arbitrarily defined biochemical cholestasis (before the collation of data) as having occurred when the serum CG was elevated either above 2 Fmol/L on at least two separate occasions or above 3 PmollL on at least one occasion during late pregnancy (after 28 wk). Using this definition we found that 266 of the 297 (90%) participants had CG concentrations remaining well within the normal range of 0-1.5 PmollL. Within this group, however, there still ex; isted a mild, but statistically significant, rise in CC from 0.3 to 0~5 pmol/L (median concentrations) during pregnancy (p < 6.65). In marked contrast, the “cholestatic” group comprising 31 (10%) subjects showed an abrupt rise in their CG concentration at
the beginning of the third trimester. This increase was maintained until delivery, after which time there was an abrupt decrease in the CG concentration (Figure 3’). We assessed the relationship between serum CG concentration and pruritus. Of the 266 subjects who maintained normal CG concentrations throughout pregnancy, 54 (20%) had significant pruritus. In contrast, 15 of the 31 participants (48%) with elevated CG admitted to significant pruritus during late pregnancy [Table 1). This difference was statistically significant (,$ = 10.8, p < 0.005). We next assessed changes in “conventional” biochemical liver tests in our two groups. The only tests
3.0
2.5‘
0 Chdastatic b Normal
2.0
Figure
I
10
15
20
25
30
weeks Pregnant
35
40
3. Changes in serum cholylglycine during pregnancy in “cholestatic” and “normal” groups (median values).
828
GASTROENTEROLOGY Vol. 91, No. 4
LUNZER ET AL.
Table 1. Relationship Between Serum Cholylglycine and Pruritus Cholylglycine concentration
No pruritus
Normal Elevated Total
212
15 227
Pruritus 54 16
Total 266 31
70
that showed abnormalities associated with elevated CG levels were the serum transaminase (aspartate transaminase and alanine transaminase) and serum cholesterol concentrations. Mild to moderate aspartate transaminase and alanine transaminase elevations (rarely exceeding three times normal) were found in 16% and 19%, respectively, of the patients with elevated CG concentrations And in only 1% and 5%, respectively, of the “normal” CG group. A raised serum cholesterol concentration was found in 42% of patients with raised CG levels and in only 20% of the normal CG group. Each of these associations was statistically significant (,$ = 20.5,8.93, and 8.01; p < 0.001, cO.005, and cO.005, respectively). No association was found between elevated CG concentratidns and serum alkaline phosphatase, y-glutamyl transferase, bilirubin, albumin, and globulin concentrations. Serum bilirubin and y-glutamyl transferage concentrations remained normal in all of the patients with elevated CG levels. No liver biopsies were performed. We studied the relationship between “biochemical cholestasis” and ethnic origin. Of the 267 patients of mid-European origin, 24 (9%) had elevated CG concentrations. In contrast, 4 of the 17 (24%) patients of Mediterranean or Asian origin had elevated CG concentrations. This difference was statistically significant (,$ = 5.94, p < 0.025). No statistically significant associations wete found between biochemical cholestasis and maternal age, number of previous pregnancies, pruritus during previous pregnancies, and fetal maturity (gestational age and birth weight). Only three participants in the study described pruritus in association with the oral contraceptive, and this symptom was not associated with pruritus dtiring pregnancy. All sera taken from patients with elevated CG levels were negative for both hepatitis B surface antigen and immunoglobulin M hepatitis A virus antibody.
Discussion Our results indicate that obstetric cholestasis may be a much more common event than previously suspected. We found significantly elevated postpran-
dial CC concentrations in 10% of our study population at 30 weeks’ pregnancy. Among this group, CG levels continued to rise significantly until delivery. Nearly one-half of this group had sustained pruritus during pregnancy. Because we found a strong association between elevated serum CG and pruritus, we believe that the majority of this group developed pruritus as a result of obstetric cholestasis. However, it should also be noted that 20% of the group maintaining normal dG levels had significant pruritus. Possible mechanisms for pruritus unrelated to cholestasis may include a hyperdynamic circulation, excessive sweating, or abdominal stretching. Mechanism$ of pruritus during pregnancy appear to be poorly understood. Thus we believe that 10% of our patients had biochemical cholestasis and nearly 5% had pruritus secondary to their cholestasis. Previous studies have underestimated the prevalence of obstetric cholestasis by only including patients with severe symptomatic disease. Although there is a well-recognized association between obstetric cholestasis and oral contraceptiveinduced cholestasis, none of our patients with elevated CG concentrations gave a history of pruritus or jaundice secondary to the oral contraceptive pill. Our data support the claim of Heikkinen (15) that the measurement of cholic acid conjugates provides the most sensitive biochemical test for the diagnosis of obstetric cholestasis. In one-half of our patients with elevated CG concentrations, mild to moderate transaminase elevations were also found but these were much less sensitive indicators of cholestasis (in the sense of strong association with pruritus) than CG concentrations. Elevated serum cholesterol concentration was also associated with elevated CG levels, but this would be a difficult test to interpret in view of the other causes of hypercholesterolemia. We have confirmed the previous study of Combes et al. (16) suggesting that y-glutamyl transferase activity may not be a useful indicator of hepatocellular disease during pregnancy. None of our patients with elevated CG concentrations had y-glutamyl transferase levels above the upper limit of normal. We have been unable to find any effect of cholestasis (clinical or subclinical) on either the gestational age or the weight of the newborn infant. It is now generally accepted that a significant number of neonates are adversely affected by obstetric cholestasis. However, we believe that this association may only apply to babies born to mothers with severe cholestasis, usually associated with jaundice, a condition not seen in any of the study subjects. In fact, serum bilirubin concentrations remained normal in all of our patients with cholestasis. It is therefore believed that we have identified a milder form of cholestasis than that usually reported in
SERUM BILE ACIDS DURING PREGNANCY
October 1986
association with deleterious effects on the newborn. Accordingly, we do not believe that our lack of association with gestational age and the weight of the newborn should be taken to indicate that more severe forms of cholestasis are not harmful to the newborn. Our study indicates that the measurement of serum bile acids may be diagnostically helpful for pregnant patients with pruritus, particularly when conventional biochemical liver tests are normal or near-normal. A normal CG concentration would make obstetric cholestasis most unlikely, but a significant elevation of CG would strongly suggest that one is dealing with pruritus secondary to cholestasis. In such instances a therapeutic trial of cholestyramine or phenobarbitone may be indicated. Both of these agents, however, have potential side-effects for both mother and fetus and should not be administered without the approval of the attending obstetrician
4. Johnston WG, Baskett TF. Obstetric cholestasis-a fourteen year review. Am J Obstet Gynecol 1979;133:299-301. 5. Laatikainen T, Ikonen E. Serum bile acids in cholestasis of pregnancy. Obstet Gynecol 1977;50:313-8. 6. Reyes H, Gonzales MC, Ribalta J. et al. Prevalence of intrahepatic cholestasis of pregnancy in Chile. Ann Intern Med 1978;88:487-93. 7. Reyes H, Taboada G, Ribalta J. Prevalence of intrahepatic cholestasis of pregnancy in La Paz. Bolivia. J Chronic Dis 1979;32:499-504. 8. Svanborg A, Ohlsson S. Recurrent jaundice of pregnancy. Am J Med 1959;27:46--9. 9. Kater RMH, Mistilis SP. Obstetric cholestasis and pruritus of pregnancy. Med J Aust 1967;1:638-40. 10. Steel R, Parker M. Jaundice in pregnancy. Med J Aust 1973; 1:461. 11. Perreau 12. 13. 14.
References 15.
Haemmerli UP, Wyss HI. Recurrent intrahepatic cholestasis of pregnancy. Medicine (Baltimore) 1967;46:299-321. Furhoff A-K, Hellstrom K. Jaundice in pregnancy. Acta Med Stand 1973;193:259-66. Reid R, Ivey KJ, Rencoret RH, et al. Fetal complications of obstetric cholestasis. Br Med J 1976;1:870-2.
829
P, Rouch R. Ictere cholestatique recidivant de la grossesse. Gynecol Obstet 1961;60:161-79. Sjovall K, Sjovall J. Serum bile acid levels in pregnancy with pruritus. Clin Chim Acta 1966;13:207-11. Fulton IC, Douglas GJ, Hutchon DJR, et al. Is normal pregnancy cholestatic? Clin Chim Acta 1983;130:170-6. Heikkinen J, Maentausta 0, Ylostalo P. et al. Changes in serum bile acid concentrations during normal pregnancy, in patients with intrahepatic cholestasis of pregnancy and in pregnant women with itching. Br J Obstet Gynaecol 1981; 88:240-5. Heikkinen J. Serum bile acids in the early diagnosis of intrahepatic cholestasis of pregnancy. Obstet Gynecol 1983:
61:581-7. 16. Combes B, Shore GM, Cunningham
FG, et al. Serum gamma glutamyl transpeptidase activity in viral hepatitis: suppression in pregnancy and by birth control pills. Gastroenterology 1977:72:271-l.
Serum Bile Acid Concentrations During Pregnancy and Their Relationship to Obstetric Cholestasis MICHAEL LUNZER, PHILIP BARNES, KAREN BYTH, and MARY O’HALLORAN Department of Gastroenterology, Royal North Shore Hospital and Department Royal Alexandra Hospital for Children, Sydney, Australia
We hove prospectively studied changes in serum postprandial cholylglycine (CG) concentration during 297 pregnancies. We found an increase in CG concentration from 0.3 mol/L at 15 weeks’ pregnancy to 0.6 PmollL at 40 weeks’ pregnancy. Although this increase was statistically significant (p < 0.005), median concentrations of CG remained well within the normal range (O-l.5 PmollL). However, 10% of the group showed markedly elevated serum CG concentrations at 30 weeks’ pregnancy, and the CC level in this group continued to rise during the third trimester. Pruritus was significantly more common in the group with elevated CG concentrations (48%) than in the group with normal CG levels (20%) (p < 0.005). Serum CG was a much more sensitive predictor of pruritus during pregnancy than other biochemical liver tests. Elevated CG levels were found more commonly in Mediterranean and Asian patients than patients of other ethnic origins [p < 0.025). No statistically signifiwere found between elevated CG cant associations concentrations and maternal age, number of previous pregnancies, pruritus during previous pregnancies, contraceptive-induced cholestasis, and fetal maturity. We conclude [a) that obstetric cholestasis is probably much more common than previously Received August 7, 1985. Accepted April 12, 1986. Address requests for reprints to: Dr. Michael Lunzer, Department of Gastroenterology, Royal North Shore Hospital, St. Leonards, Sydney, N.S.W. 2065, Australia. The authors thank Abbott Diagnostics, North Chicago, Illinois for their generous donation of radioimmunoassay kits for the assessment of serum cholylglycine concentration. The authors thank Elaine Brecknock and Leslie McAdam for collection of data, Dr. Hilary Lunzer for computation of data, and Evelyn Humpherson for manuscript preparation. They also acknowledge the cooperation of the Medical Staff of the Department of Obstetrics at Royal North Shore Hospital and thank them for allowing them to study their patients. 8 1986 by the American Gastroenterological Association 0016-5085/86/$3.50
of Biochemistry,
suspected and (b) that consideration should be given to the measurement of serum bile acids in all pregnant individuals with unexplained pruritus. Obstetric cholestasis is an unusual condition characterized by pruritus and, occasionally, jaundice during the latter part of pregnancy. The symptoms and biochemical abnormalities of cholestasis typically subside soon after birth and tend to recur during subsequent pregnancies (1).An association exists between obstetric cholestasis and oral contraceptive-induced cholestasis, and the condition may be familial (2). Although obstetric cholestasis does not have a deleterious effect on the mother, there are several reports of significant risk to the fetus including prematurity, low birth weight, intrapartum asphyxia, and increased fetal mortality (3-5). Obstetric cholestasis is believed to be a rare condition outside certain Scandinavian and South American countries. Reported prevalence rates for obstetric cholestasis in Chile (12%22%) (6), Bolivia (9%) (7), and Sweden (2%3%) (8) are considerably higher than those found in Australia (O-2%-0.8%) (9,10), France (0.2%) (ll), and Canada (0.1%)(4). No prevalence data are available for the United States, Britain, or most other parts of the world. Mild abnormalities of the routine biochemical liver tests are frequently found in patients with significant obstetric cholestasis. The most marked abnormalities, however, have been reported in the serum concentration of bile acids. Sjovall and Sjovall (12) found 30-fold elevation of serum total bile acids in 6 patients with obstetric cholestasis. The most abnormal elevation of the individual bile acids was that of cholic acid, which was elevated i’O-fold. Similar results were reported by Laatikainen and Ikonen (5). Fulton et al. (l3), in a small series of Abbreviation
used in this paper: CG, cholylglycine.
826
GASTROENTEROLOGY Vol. 91. No. 4
LUNZER ET AL.
26 women, found elevated fasting (but not postprandial) concentrations of cholate after 34 weeks’ gestation. Heikkinen et al. (14) found lo-fold elevations in serum levels of cholic acid in women with obstetric cholestasis, and noted that in 4 of 8 women elevation of serum bile acid levels preceded the appearance of symptoms or other laboratory evidence of intrahepatic cholestasis. There have, however, been no previous reports of a large population on whom serial measurements of serum bile acids have been performed prospectively during pregnancy. In the present study we have chosen to use a commercially available assay for estimating the glycine conjugate of cholic acid. Although the assay measures predominantly the glycine conjugate, the antibody does have cross-reactivity with taurocholic acid (16.5%) and glycochenodeoxycholic acid (14.5%). Some investigators have claimed that the sulfate esters of bile acids represent the major form in hepatobiliary disease; however, the main component of this sulfated group, sulfochenodeoxycholic acid, is not easily assayed in “routine” laboratories. It should also be pointed out that a good correlation exists between the concentration of sulfated bile acids and the concentration of serum glycocholate. The aim of the present study was to define the changes in serum cholylglycine (CG) concentration in pregnant women and to note the relationship of these changes to pruritus and other clinical and biochemical parameters.
Materials and Methods Two hundred ninety-seven pregnant individuals attending the Ante-Natal Clinic at Royal North Shore Hospital were prospectively studied during a 21-mo period from March 1981 to December 1982. All patients attending the clinic during this period were given an opportunity to participate in the study; the 297 participants represented those individuals who agreed to participate after being provided with full details of the requirements of the study. The mean age of participants was 27 yr (range 16-42 yr) and the mean number of previous pregnancies was 1.1. Participants were seen on at least five occasions during pregnancy. The initial visit was at -15 weeks’ gestation, with subsequent visits at -26, 32, 38, and 38 weeks’ gestation. At each visit patients were asked about symptoms of cholestasis (pruritus, dark urine, pale stools, jaundice) and drug ingestion before and during the current pregnancy. Participants were considered to have sustained pruritus if this symptom was present on at least two occasions during late pregnancy (after 28 weeks’ gestation). There were no patients with known hepatobiliary, intestinal, or pancreatic disease in the study. No patients gave a history of alcohol consumption exceeding 40 g daily. No patients were taking drugs known to affect hepatobiliary function. The ethnic origin of participants was also ascertained.
fn i; F
30.
g0 8
II 20.
I3 g L
I --
MEDIAN
= 0.5 JIM
RANGE
=O.l
(N=50)
10.
1
2
1. Two-hour postprandial tions in controls.
4
3
SERUM CHOLYLGLYCINE
Figure
- 1.5 JJM
~JM
serum cholylglycine
concentra-
Participants were asked to have a meal containing a standard amount of fat (18 g) 2 h before their visit, and blood was taken at each visit for assay of serum CG and conventional biochemical liver tests (bilirubin, alkaline phosphatase, -y-glutamyl transferase, aspartate transaminase, alanine transaminase, albumin, globulins). It was not considered justifiable to take multiple samples for serum CG measurement at clinic visits. Blood was also taken for the above tests 48-72 h postpartum. Serum CG was measured by radioimmunoassay (Abbott Diagnostics, North Chicago, Ill.) and “conventional” biochemical liver tests were measured by a S.M.A.C. autoanalyzer. Serum was assayed for hepatitis B surface antigen and immunoglobulin M hepatitis A virus antibody by radioimmunoassay in all patients with elevated serum CG concentrations on sera that showed the highest CG concentration. The computer package S.P.S.S. (Release 8) was used to analyze the data. The 2 and paired t-tests were used, as were linear regression techniques utilizing the F test for linear effect allowing for possible different intercepts for individuals.
Results A control population of 50 age-matched, nonpregnant female volunteers was obtained and data for 2-h postprandial serum CG concentrations in this group are shown in Figure 1. A normal range of o-l.5 pmol/L was found. The changes in serum CG found during pregnancy are shown in Figure 2. Serum CG rose progressively from a median concentration of 0.3 PmollL in early pregnancy to 0.7 pmol/L in late pregnancy, with a fall to 0.6 pmol/L in the immediate postpartum period. Although the rise in CG remained well within the normal range, the increase was statistically significant (p < 0.005). The fall in CG immediately after delivery also reached statistical significance (p < 0.01)
October 1986
SERUM BILE ACIDS DURING PREGNANCY
827
1.5 -
1.2%
1.0 -
Figure
n-t -10
15
2b
25
30
3j
2. Changes in serum cholylglycine during pregnancy (median values].
40
Weeks Pregnant
We arbitrarily defined biochemical cholestasis (before the collation of data) as having occurred when the serum CG was elevated either above 2 Fmol/L on at least two separate occasions or above 3 PmollL on at least one occasion during late pregnancy (after 28 wk). Using this definition we found that 266 of the 297 (90%) participants had CG concentrations remaining well within the normal range of 0-1.5 PmollL. Within this group, however, there still ex; isted a mild, but statistically significant, rise in CC from 0.3 to 0~5 pmol/L (median concentrations) during pregnancy (p < 6.65). In marked contrast, the “cholestatic” group comprising 31 (10%) subjects showed an abrupt rise in their CG concentration at
the beginning of the third trimester. This increase was maintained until delivery, after which time there was an abrupt decrease in the CG concentration (Figure 3’). We assessed the relationship between serum CG concentration and pruritus. Of the 266 subjects who maintained normal CG concentrations throughout pregnancy, 54 (20%) had significant pruritus. In contrast, 15 of the 31 participants (48%) with elevated CG admitted to significant pruritus during late pregnancy [Table 1). This difference was statistically significant (,$ = 10.8, p < 0.005). We next assessed changes in “conventional” biochemical liver tests in our two groups. The only tests
3.0
2.5‘
0 Chdastatic b Normal
2.0
Figure
I
10
15
20
25
30
weeks Pregnant
35
40
3. Changes in serum cholylglycine during pregnancy in “cholestatic” and “normal” groups (median values).
828
GASTROENTEROLOGY Vol. 91, No. 4
LUNZER ET AL.
Table 1. Relationship Between Serum Cholylglycine and Pruritus Cholylglycine concentration
No pruritus
Normal Elevated Total
212
15 227
Pruritus 54 16
Total 266 31
70
that showed abnormalities associated with elevated CG levels were the serum transaminase (aspartate transaminase and alanine transaminase) and serum cholesterol concentrations. Mild to moderate aspartate transaminase and alanine transaminase elevations (rarely exceeding three times normal) were found in 16% and 19%, respectively, of the patients with elevated CG concentrations And in only 1% and 5%, respectively, of the “normal” CG group. A raised serum cholesterol concentration was found in 42% of patients with raised CG levels and in only 20% of the normal CG group. Each of these associations was statistically significant (,$ = 20.5,8.93, and 8.01; p < 0.001, cO.005, and cO.005, respectively). No association was found between elevated CG concentratidns and serum alkaline phosphatase, y-glutamyl transferase, bilirubin, albumin, and globulin concentrations. Serum bilirubin and y-glutamyl transferage concentrations remained normal in all of the patients with elevated CG levels. No liver biopsies were performed. We studied the relationship between “biochemical cholestasis” and ethnic origin. Of the 267 patients of mid-European origin, 24 (9%) had elevated CG concentrations. In contrast, 4 of the 17 (24%) patients of Mediterranean or Asian origin had elevated CG concentrations. This difference was statistically significant (,$ = 5.94, p < 0.025). No statistically significant associations wete found between biochemical cholestasis and maternal age, number of previous pregnancies, pruritus during previous pregnancies, and fetal maturity (gestational age and birth weight). Only three participants in the study described pruritus in association with the oral contraceptive, and this symptom was not associated with pruritus dtiring pregnancy. All sera taken from patients with elevated CG levels were negative for both hepatitis B surface antigen and immunoglobulin M hepatitis A virus antibody.
Discussion Our results indicate that obstetric cholestasis may be a much more common event than previously suspected. We found significantly elevated postpran-
dial CC concentrations in 10% of our study population at 30 weeks’ pregnancy. Among this group, CG levels continued to rise significantly until delivery. Nearly one-half of this group had sustained pruritus during pregnancy. Because we found a strong association between elevated serum CG and pruritus, we believe that the majority of this group developed pruritus as a result of obstetric cholestasis. However, it should also be noted that 20% of the group maintaining normal dG levels had significant pruritus. Possible mechanisms for pruritus unrelated to cholestasis may include a hyperdynamic circulation, excessive sweating, or abdominal stretching. Mechanism$ of pruritus during pregnancy appear to be poorly understood. Thus we believe that 10% of our patients had biochemical cholestasis and nearly 5% had pruritus secondary to their cholestasis. Previous studies have underestimated the prevalence of obstetric cholestasis by only including patients with severe symptomatic disease. Although there is a well-recognized association between obstetric cholestasis and oral contraceptiveinduced cholestasis, none of our patients with elevated CG concentrations gave a history of pruritus or jaundice secondary to the oral contraceptive pill. Our data support the claim of Heikkinen (15) that the measurement of cholic acid conjugates provides the most sensitive biochemical test for the diagnosis of obstetric cholestasis. In one-half of our patients with elevated CG concentrations, mild to moderate transaminase elevations were also found but these were much less sensitive indicators of cholestasis (in the sense of strong association with pruritus) than CG concentrations. Elevated serum cholesterol concentration was also associated with elevated CG levels, but this would be a difficult test to interpret in view of the other causes of hypercholesterolemia. We have confirmed the previous study of Combes et al. (16) suggesting that y-glutamyl transferase activity may not be a useful indicator of hepatocellular disease during pregnancy. None of our patients with elevated CG concentrations had y-glutamyl transferase levels above the upper limit of normal. We have been unable to find any effect of cholestasis (clinical or subclinical) on either the gestational age or the weight of the newborn infant. It is now generally accepted that a significant number of neonates are adversely affected by obstetric cholestasis. However, we believe that this association may only apply to babies born to mothers with severe cholestasis, usually associated with jaundice, a condition not seen in any of the study subjects. In fact, serum bilirubin concentrations remained normal in all of our patients with cholestasis. It is therefore believed that we have identified a milder form of cholestasis than that usually reported in
SERUM BILE ACIDS DURING PREGNANCY
October 1986
association with deleterious effects on the newborn. Accordingly, we do not believe that our lack of association with gestational age and the weight of the newborn should be taken to indicate that more severe forms of cholestasis are not harmful to the newborn. Our study indicates that the measurement of serum bile acids may be diagnostically helpful for pregnant patients with pruritus, particularly when conventional biochemical liver tests are normal or near-normal. A normal CG concentration would make obstetric cholestasis most unlikely, but a significant elevation of CG would strongly suggest that one is dealing with pruritus secondary to cholestasis. In such instances a therapeutic trial of cholestyramine or phenobarbitone may be indicated. Both of these agents, however, have potential side-effects for both mother and fetus and should not be administered without the approval of the attending obstetrician
4. Johnston WG, Baskett TF. Obstetric cholestasis-a fourteen year review. Am J Obstet Gynecol 1979;133:299-301. 5. Laatikainen T, Ikonen E. Serum bile acids in cholestasis of pregnancy. Obstet Gynecol 1977;50:313-8. 6. Reyes H, Gonzales MC, Ribalta J. et al. Prevalence of intrahepatic cholestasis of pregnancy in Chile. Ann Intern Med 1978;88:487-93. 7. Reyes H, Taboada G, Ribalta J. Prevalence of intrahepatic cholestasis of pregnancy in La Paz. Bolivia. J Chronic Dis 1979;32:499-504. 8. Svanborg A, Ohlsson S. Recurrent jaundice of pregnancy. Am J Med 1959;27:46--9. 9. Kater RMH, Mistilis SP. Obstetric cholestasis and pruritus of pregnancy. Med J Aust 1967;1:638-40. 10. Steel R, Parker M. Jaundice in pregnancy. Med J Aust 1973; 1:461. 11. Perreau 12. 13. 14.
References 15.
Haemmerli UP, Wyss HI. Recurrent intrahepatic cholestasis of pregnancy. Medicine (Baltimore) 1967;46:299-321. Furhoff A-K, Hellstrom K. Jaundice in pregnancy. Acta Med Stand 1973;193:259-66. Reid R, Ivey KJ, Rencoret RH, et al. Fetal complications of obstetric cholestasis. Br Med J 1976;1:870-2.
829
P, Rouch R. Ictere cholestatique recidivant de la grossesse. Gynecol Obstet 1961;60:161-79. Sjovall K, Sjovall J. Serum bile acid levels in pregnancy with pruritus. Clin Chim Acta 1966;13:207-11. Fulton IC, Douglas GJ, Hutchon DJR, et al. Is normal pregnancy cholestatic? Clin Chim Acta 1983;130:170-6. Heikkinen J, Maentausta 0, Ylostalo P. et al. Changes in serum bile acid concentrations during normal pregnancy, in patients with intrahepatic cholestasis of pregnancy and in pregnant women with itching. Br J Obstet Gynaecol 1981; 88:240-5. Heikkinen J. Serum bile acids in the early diagnosis of intrahepatic cholestasis of pregnancy. Obstet Gynecol 1983:
61:581-7. 16. Combes B, Shore GM, Cunningham
FG, et al. Serum gamma glutamyl transpeptidase activity in viral hepatitis: suppression in pregnancy and by birth control pills. Gastroenterology 1977:72:271-l.