Original Paper
Digestion 1992;52:222-231
Division o f Gastroenterology, Department of Medicine, Tri-Service General Hospital, and School of Pharmacy, National Defense Medical Center, Taipei, Republic of China
Assessment of Liver Function Using a Novel Galactose Single Point Method
Keywords
Abstract
Galactose single point method Modified galactose elimination capacity Galactose elimination capacity Residual liver function Chronic hepatitis Cirrhosis Hepatocellular carcinoma
A novel, simple, clinically useful quantitative liver function test, called the galactose single point (GSP) method, was developed by measurement of galactose blood concentration 1 h after galactose was administered (0.5 g/kg). It was quickly infused intravenously in 55 normal healthy volun teers, 73 patients with chronic hepatitis (CH), 36 with cirrhosis and 41 with hepatocellular carcinoma (HCC). Patients with CH diagnosis were assessed by liver biopsy. Cirrhosis was diagnosed by histological examina tion or a chronic hepatitis history with esophageal varices or ascites, whereas HCC was diagnosed either histologically, or cytologically proved, or as implied in the ‘one imagine study’ being positive with AFP > 300 ng/dl. Highly significant galactose blood levels were observed between normal healthy volunteers and patients 50. 60 and 70 min after galactose was administered. Galactose elimination capacity (GEC), modified GEC (MGEC) and consecutive GSP tests were performed in 6 healthy volun teers for 2 days. 0.64-16.87 % variation was observed for each subject. The significant differences (p < 0.001) in average GSP values were 247 ± 18.1.422 ± 27.3, 629 ± 42.8 and 579 ± 43.6 pg/ml for normal healthy volunteers, CH, cirrhosis and HCC patients, respectively. Highly signifi cant correlations (p < 0.001) were obtained among GSP. GEC and MGEC for all patients. Positive correlations were observed between GSP, GEC, MGEC and AST (serum aspartate aminotransferase), ALT (serum alanine aminotransferase), serum bilirubin, albumin, prothrombin time and r-globulin. According to results obtained from 202 normal healthy volunteers and patients, the GSP method may be a simple, clinically useful quantitative measurement of liver function for the determination of a patient’s residual liver function, the prognosis of liver function for patients with cirrhosis, postoperational follow-up and, finally, the timing of a liver transplant.
Received: January 13,1992 Received in revised form: June 22.1992
Dr. Hung-Shang Tang. MD Division o f Gastroenterology. Department of Medicine Tri-Service General Hospital National Defense Medical Center. 10.3 section Ting-Chow Road. Taipei. Taiwan (Republic of China)
©1992 S. Karger AG, Basel 0012-2823/92/ 0524-0222S2.75/0
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/5/2018 10:13:40 AM
Hung-Shang Tang a Oliver Yoa-Pu H ub
mals [16] and humans [17], galactose follows the Michaelis-Menten saturation kinetics. Therefore, Tygstrup [ 18] developed the galac tose elimination capacity (GEC) method to quantitatively estimate the liver function, i.e. intrinsic hepatic clearance from the slope of the galactose concentration time plot 2060 min after galactose was administered. However, this method needs at least four blood samples for estimating the value of GEC. Also, the 20- to 60-min portion of the galactose blood concentration time curve is not always the saturated portion for normal healthy subjects and different patients. The first report of these series of investiga tions deals with the development of a simple, clinically useful method called the galactose single point (GSP) method, which can quanti tatively measure residual liver function and correlates reasonably well with the GEC, clin ical biochemical and histological change in patients with chronic hepatitis (CH). cirrhosis and hepatocellular carcinoma (HCC). Material and Methods Subjects Fifty-five normal healthy volunteers, aged 19-35 years, were selected to participate in this study based on the results of medical history, physical and clinical examinations and laboratory diagnosis including blood chemistry, urine analysis and SMA-12. During the years 1987-1989. patients between the age of 17 and 76 with CH (n = 73). cirrhosis (n = 36) and HCC (n = 41) were selected from Tri-Service General Hospi tal. Patients with CH diagnosis were assessed by liver biopsy, whereas cirrhosis was diagnosed by histologi cal examination or CH history with esophageal varices or ascites. According to the Child-Pugh classification. 16 cirrhosis patients were classified as A, 12 as B, and the rest In = 8) was classified as C [ 19], HCC was diag nosed either histologically or cytologically as implied in the "one imagine study’ being positive with AFP > 3 0 0 ng/dl. Patients with diabetes mellitus were ex cluded. The study was approved by the Institutional Review Board o f the National Defense Medical Cen ter.
223
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/5/2018 10:13:40 AM
Currently, quantitative measurement of residual liver function is not a routine clinical test. Quantitative measurement of the entire liver function including all processes which take place in the liver is not possible funda mentally [1, 2], However, quantitative mea surement of kidney function is routinely per formed in clinics by measuring serum creati nine concentration, based on the assumption that the degree of kidney function, i.e. glomer ular filtration, active tubular secretion and tubular reabsorption, is proportional to the function of glomerular filtration [3]. Measure ment of the metabolic capacity of a certain process can assume a metabolic rate-limiting step in the liver, making it possible to obtain a representative value of residual liver func tion. A number of methods and drug clearances were used to estimate liver functions. Exam ples for hepatic blood flow include: high ex traction compounds, lidocaine [4] and pro pranolol [5], Methods used by the intrinsic clearance were low extraction compounds, antipyrine [6], aminopyrine [7], diazepam [8], chlordiazepoxide [9] and caffeine [10], These drugs may increase the burden of the liver which has already suffered from a certain degree of damage. Galactose is a naturally occurring sugar with a high extraction ratio that is 90% me tabolized in the liver. In the liver, galactose is catalyzed by galactokinase through epimerization, thus converting to glucose-1-phos phate. The reaction with galactokinase is the rate-limiting step of the galactose metabolism in the liver [11-14], Since galactose is a high extraction ratio compound, based on the ‘well stir model’, the clearance of the first order process at a low concentration can represent the hepatic blood flow [ 15], When a large amount of galactose is administered, galac tose metabolism can be saturated due to the limited capacity of galactokinase [12], In ani
Galactose Injection Galactose injection solution was prepared by the Kingdom Pharmaceutical Company in Taipei. Repub lic of China, using 400 g galactose (Sigma. USA) dis solved in 1.0 liter distilled water for injection. Sterility, pyrogen tests and other general injection tests were performed according to the USPXXI.
Measurement o f Galactose Blood Concentration The galactose blood concentration was measured by an enzymatic method [ 17], 50 pi of blood were col lected. using disposable microliter pipettes, then added to 0.5 ml EDO in a test tube kept on ice. 100 pi of 15% perchloric acid was then added, to precipitate protein. After neutralization. NADPH, hydrazine and galac tose dehydrogenase were added; galactose concentra tion was determined by fluorescent spectroscopy (exci tation: 365 nm, emission: 472 nni). The lower limit of the working range was 49.82 pg/ml. which was linear with concentrations up to 1,901 pg/ml. The galactose calibration curve was constructed for each subject using his own blank blood withdrawn before galactose was administered. Each patient’s en dogenous galactose was measured and deducted from each serum sample. Duplicate galactose blood concen trations. 49.82 99.30. 482.8. 933.3. 1.355 and 1,901 pg/ml were spiked and assayed. After linear regression of the mean, the calibration curve was constructed. Repeatability. The calibration curve was con structed using the subject’s own blood each day. Thus
224
Tang/Hu
0
20
40
60
80
1 00
120
140
1 60
160
200
Time, min
Fig. 1. The mean venous blood concentration-time profile of galactose after quick intravenous injection of 0.5 g/kg galactose in 55 normal healthy volunteers and 36 patients with cirrhosis.
the slope, intercept and percent CV of each calibration curve was monitored. Six healthy subjects received galactose independently every morning which was re peated on consecutive days. The GSP. GEC and the modified GEC MGEC values were recorded and com pared as this activity occurred. GEC. GEC was calculated according to the follow ing formula, which was modified from Tygstrup’s [20] equation GEC = — —— (mg/kg-min). tc . o + 7 where D is the injected amount of galactose. tc.o is the amount of time required for the concentration to reach zero, extrapolated from a linear regression of the blood concentration-time curse from 20 min after injection to 60 min (usually, the galactose concentration at 60 min was higher than 2.22 mmol/1). and 7 is the empirical correction for uneven distribution in the body. MGEC. Galactose disposition in humans follows two pharmacokinetic open models: one rapid distribu tion phase followed by an elimination phase (fig. I ). Under the saturated dose of 0.5 g/kg. the linear portion (saturated portion) of galactose blood level in normal humans and in patients was different (fig. 1.2). Signifi cant variations in the distribution phase and the satu-
The Novel Liver Function Test GSP
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/5/2018 10:13:40 AM
Study Design All patients and volunteers underwent 3 min of quick intravenous administration of 0.5 g/kg galactose solution (0.4 g/nil). A butterfly needle was placed in a forearm vein. Volunteers and patients were kept in bed to rest for 3 h. The galactose tests consistently began at 9 a.m. and were completed by 12 a.m. The GEC was assessed largely according to Tygstrup’s [ 18] procedure [12. 16] using venous blood samples (0.5 ml) drawn at 5. 10. 15, 20, 25. 30. 35, 40, 50. 60. 70, 80, 90. 100, 110, 120, 130, 140. 150. 160, 170 and 180 min after the injection. Blood samples were kept in the ice bath until measured by an enzymatic method [17], A dupli cate. six-concentration calibration curve, ranging from 49.82 to 1.90! pg/ml, was constructed using blank blood drawn before administering to each subject. Within-day variation was evaluated by standard de viation and percent coefficient of variation (CV) for each concentration. A maximal of 10% CV was per mitted. Day-to-day variation was also checked by com paring the slope and intercept of the calibration curves.
Time, min
Time, min
3 26
5
10 20 2 5 3 0 40 5 0 00 70 SO 9 0 100
Time, min
2
rated portion among volunteers and patients were also noted (fig. 2). The rationale for the estimation of tc =o by extrapolating the regression line from 20 to 60 min was based on the assumption that the saturated por tion of the galactose-time curve was between 20 and 60 min. MGEC was estimated using each person’s linear portion of galactose concentration-time curve which is the real saturated portion after the distribution phase. GSP Method. After observing 52 normal healthy volunteers and 36 patients with hepatic cirrhosis in the MGEC study, highly significant different galactose blood levels were found between healthy people and patients at 40, 50 and 60 min following galactose administration (fig. 3). The galactose blood concentration was measured I h after administration and was selected as the GSP which measured the quantitative liver function.
Statistical Analysis Statistical evaluation utilized the analysis of vari ance (Anova) followed by the multiple range test, least significant difference (LSD), linear regression. Stu dent’s t test and Pearson's correlation coefficient test.
Results The sex. age and the possible etiology are presented in table 1. One normal subject, 2 with cirrhosis and 2 HCC patients did not complete the study. Results show that virus hepatitis is the major cause of liver dysfunc tion in Taiwan. The precision of the galactose
225
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/5/2018 10:13:40 AM
0
Fig. 2. The galactose saturation time period in nor mal volunteers (a) and patients with cirrhosis (b) obtained from each person’s galactose concentration time curve after receiving quick intravenous adminis tration of 0.5 g/kg galactose. Fig. 3. Difference of galactose blood concentration at different time points between 51 normal healthy volunteers and 36 patients with cirrhosis after 0.5 g/kg galactose was infused intravenously for 3 min. The concentrations 40. 50 and 60 min after infusion give highly significant differences between these two groups.
Table 1. Sex. age and/or the possible etiology of normal subjects and patients with CH. cirrhosis and HCC
Normal CH Cirrhosis HCC
Sex M
F
mean ± SD
range
HBV
HCV
alcoholic others
51 73 34 39
4 0 2 2
23.4 ±3.6 25.6 ±9.2 40.8 ±17.6 50.8 ±14.8
19-35 17-68 19-67 32-76
0 65 26 29
0 5 6 4
0 0 2 5
calibration curve was excellent with less than 6.66% within-day correlation coefficient (n = 3). The slope of each between-day calibration curve was compared, and the mean and stan dard derivations of slopes were recorded and monitored (fig. 4). Over 90% of the slopes fall in the range of the mean ± 2 SD. The mea sured and added galactose concentrations were also monitored for the accuracy of the assay. The estimated repetitive GSP. GEC and MGEC values for 6 healthy subjects who received galactose independently every morn ing on 2 consecutive days are listed in table 2. The CV for GEC. MGEC and GSP ranged from 0.64 to 16.87%. Table 3 shows the results of GSP, GEC and MGEC for normal healthy volunteers, pa tients with CH. cirrhosis and HCC. Highly significant differences were observed among these subjects (Anova, p < 0.001). Similar results were obtained using GEC and MGEC methods. Patients with HCC and cirrhosis were not significantly different (p > 0.05: using GSP: also there was no highly sig nificant difference (p < 0.05 only) using MGEC and GEC (table 3). Highly significant GSP values were observed among normal subjects, patients with CH, cirrhosis and HCC (p < 0.001). However, patients with CH and HCC showed no highly significant differ ences (p < 0.01 only) using GEC. This was
226
Possible etiology
Age, years
Tang/Hu
0 3 2 3
Table 2. The repeated estimated GSP, GEC and MGEC values on 2 consecutive days in 6 healthy sub jects
Subject
GSP pg/ml (% CV)
GEC mg/ml-kg (% CV)
MGEC mg/mbkg (% CV)
1
388-464 (8.92%) 111.4-156.6 (16.87%) 259.8-328.2 (11.63%) 129.1-178.9 (16.17%) 124.7-151.3 (9.64%) 228.74-241.26 (2.66%)
5.12-5.78 (6.06%) 6.97-7.41 (3.06%) 5.99-6.55 (4.42%) 6.97-7.37 (4.1%) 6.96-7.32 (2.52%) 6.21-6.29 (2.64%)
5.12-5.78 (5.97%) 7.47-8.97 (9.12%) 5.53-6.07 (4.62%) 7.25-7.73 (3.24%) 5.80-6.52 (5.38%) 6.03-6.35 (2.62%)
2 3 4 5 6
% CV was calculated using the standard deviation on 2 consecutive days of GSP. GEC or MGEC values divided by their respective means.
even worse while using MGEC. Nonsignifi cant results were obtained between patients with CH and HCC. GSP had highly significant correlations with both GEC (fig. 5) and MGEC (p < 0.001) with the correlation coefficients being 0.909 and 0.788. respectively: MGEC also
The Novel l iver Function Test GSP
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/5/2018 10:13:40 AM
Subjects
Fig. 4. Slope (mean ± SD and mean ± 2SD) of calibration curves from 41 subjects' own blank whole blood.
Fig. 5. The correlation between GEC and GSP among 173 healthy subjects and patients with CH. cir rhosis and HCC.
Table 3. Values and statistical results of GSP, GEC and MGEC (means ± SE) in 46 normal subjects (N). 66 patients with CH. 33 cirrhosis (C) and 36 HCC after 0.5 g/kg galactose was intravenously administered
Normal
CH
Cirr hosis
HCC
Statistical analysis: Anova & LSD N-CH
N-C
N-HCC
CH-C
CH-HCC C-HCC
GSP pg/ml-kg
247.2 ±16.5
423 ±26.0
630 ±41.0
580 ±40.9
< 0 .0 1
< 0.001
< 0.001
< 0.001
< 0.001
NS
GEC mg/ml •kg
6.68 ±0.07
5.80 ±0.11
4.78 ±0.17
5.28 ±0.18
< 0 .0 1
< 0.001
< 0.001
< 0.0 0 1
< 0 .0 1
< 0.05
MGEC mg/ml-kg
7.18 ±0.15
5.75 ±0.15
4.72 ±0.23
5.49 ±0.27
< 0 .0 1
Digestion 1992;52:222-231
Division o f Gastroenterology, Department of Medicine, Tri-Service General Hospital, and School of Pharmacy, National Defense Medical Center, Taipei, Republic of China
Assessment of Liver Function Using a Novel Galactose Single Point Method
Keywords
Abstract
Galactose single point method Modified galactose elimination capacity Galactose elimination capacity Residual liver function Chronic hepatitis Cirrhosis Hepatocellular carcinoma
A novel, simple, clinically useful quantitative liver function test, called the galactose single point (GSP) method, was developed by measurement of galactose blood concentration 1 h after galactose was administered (0.5 g/kg). It was quickly infused intravenously in 55 normal healthy volun teers, 73 patients with chronic hepatitis (CH), 36 with cirrhosis and 41 with hepatocellular carcinoma (HCC). Patients with CH diagnosis were assessed by liver biopsy. Cirrhosis was diagnosed by histological examina tion or a chronic hepatitis history with esophageal varices or ascites, whereas HCC was diagnosed either histologically, or cytologically proved, or as implied in the ‘one imagine study’ being positive with AFP > 300 ng/dl. Highly significant galactose blood levels were observed between normal healthy volunteers and patients 50. 60 and 70 min after galactose was administered. Galactose elimination capacity (GEC), modified GEC (MGEC) and consecutive GSP tests were performed in 6 healthy volun teers for 2 days. 0.64-16.87 % variation was observed for each subject. The significant differences (p < 0.001) in average GSP values were 247 ± 18.1.422 ± 27.3, 629 ± 42.8 and 579 ± 43.6 pg/ml for normal healthy volunteers, CH, cirrhosis and HCC patients, respectively. Highly signifi cant correlations (p < 0.001) were obtained among GSP. GEC and MGEC for all patients. Positive correlations were observed between GSP, GEC, MGEC and AST (serum aspartate aminotransferase), ALT (serum alanine aminotransferase), serum bilirubin, albumin, prothrombin time and r-globulin. According to results obtained from 202 normal healthy volunteers and patients, the GSP method may be a simple, clinically useful quantitative measurement of liver function for the determination of a patient’s residual liver function, the prognosis of liver function for patients with cirrhosis, postoperational follow-up and, finally, the timing of a liver transplant.
Received: January 13,1992 Received in revised form: June 22.1992
Dr. Hung-Shang Tang. MD Division o f Gastroenterology. Department of Medicine Tri-Service General Hospital National Defense Medical Center. 10.3 section Ting-Chow Road. Taipei. Taiwan (Republic of China)
©1992 S. Karger AG, Basel 0012-2823/92/ 0524-0222S2.75/0
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/5/2018 10:13:40 AM
Hung-Shang Tang a Oliver Yoa-Pu H ub
mals [16] and humans [17], galactose follows the Michaelis-Menten saturation kinetics. Therefore, Tygstrup [ 18] developed the galac tose elimination capacity (GEC) method to quantitatively estimate the liver function, i.e. intrinsic hepatic clearance from the slope of the galactose concentration time plot 2060 min after galactose was administered. However, this method needs at least four blood samples for estimating the value of GEC. Also, the 20- to 60-min portion of the galactose blood concentration time curve is not always the saturated portion for normal healthy subjects and different patients. The first report of these series of investiga tions deals with the development of a simple, clinically useful method called the galactose single point (GSP) method, which can quanti tatively measure residual liver function and correlates reasonably well with the GEC, clin ical biochemical and histological change in patients with chronic hepatitis (CH). cirrhosis and hepatocellular carcinoma (HCC). Material and Methods Subjects Fifty-five normal healthy volunteers, aged 19-35 years, were selected to participate in this study based on the results of medical history, physical and clinical examinations and laboratory diagnosis including blood chemistry, urine analysis and SMA-12. During the years 1987-1989. patients between the age of 17 and 76 with CH (n = 73). cirrhosis (n = 36) and HCC (n = 41) were selected from Tri-Service General Hospi tal. Patients with CH diagnosis were assessed by liver biopsy, whereas cirrhosis was diagnosed by histologi cal examination or CH history with esophageal varices or ascites. According to the Child-Pugh classification. 16 cirrhosis patients were classified as A, 12 as B, and the rest In = 8) was classified as C [ 19], HCC was diag nosed either histologically or cytologically as implied in the "one imagine study’ being positive with AFP > 3 0 0 ng/dl. Patients with diabetes mellitus were ex cluded. The study was approved by the Institutional Review Board o f the National Defense Medical Cen ter.
223
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/5/2018 10:13:40 AM
Currently, quantitative measurement of residual liver function is not a routine clinical test. Quantitative measurement of the entire liver function including all processes which take place in the liver is not possible funda mentally [1, 2], However, quantitative mea surement of kidney function is routinely per formed in clinics by measuring serum creati nine concentration, based on the assumption that the degree of kidney function, i.e. glomer ular filtration, active tubular secretion and tubular reabsorption, is proportional to the function of glomerular filtration [3]. Measure ment of the metabolic capacity of a certain process can assume a metabolic rate-limiting step in the liver, making it possible to obtain a representative value of residual liver func tion. A number of methods and drug clearances were used to estimate liver functions. Exam ples for hepatic blood flow include: high ex traction compounds, lidocaine [4] and pro pranolol [5], Methods used by the intrinsic clearance were low extraction compounds, antipyrine [6], aminopyrine [7], diazepam [8], chlordiazepoxide [9] and caffeine [10], These drugs may increase the burden of the liver which has already suffered from a certain degree of damage. Galactose is a naturally occurring sugar with a high extraction ratio that is 90% me tabolized in the liver. In the liver, galactose is catalyzed by galactokinase through epimerization, thus converting to glucose-1-phos phate. The reaction with galactokinase is the rate-limiting step of the galactose metabolism in the liver [11-14], Since galactose is a high extraction ratio compound, based on the ‘well stir model’, the clearance of the first order process at a low concentration can represent the hepatic blood flow [ 15], When a large amount of galactose is administered, galac tose metabolism can be saturated due to the limited capacity of galactokinase [12], In ani
Galactose Injection Galactose injection solution was prepared by the Kingdom Pharmaceutical Company in Taipei. Repub lic of China, using 400 g galactose (Sigma. USA) dis solved in 1.0 liter distilled water for injection. Sterility, pyrogen tests and other general injection tests were performed according to the USPXXI.
Measurement o f Galactose Blood Concentration The galactose blood concentration was measured by an enzymatic method [ 17], 50 pi of blood were col lected. using disposable microliter pipettes, then added to 0.5 ml EDO in a test tube kept on ice. 100 pi of 15% perchloric acid was then added, to precipitate protein. After neutralization. NADPH, hydrazine and galac tose dehydrogenase were added; galactose concentra tion was determined by fluorescent spectroscopy (exci tation: 365 nm, emission: 472 nni). The lower limit of the working range was 49.82 pg/ml. which was linear with concentrations up to 1,901 pg/ml. The galactose calibration curve was constructed for each subject using his own blank blood withdrawn before galactose was administered. Each patient’s en dogenous galactose was measured and deducted from each serum sample. Duplicate galactose blood concen trations. 49.82 99.30. 482.8. 933.3. 1.355 and 1,901 pg/ml were spiked and assayed. After linear regression of the mean, the calibration curve was constructed. Repeatability. The calibration curve was con structed using the subject’s own blood each day. Thus
224
Tang/Hu
0
20
40
60
80
1 00
120
140
1 60
160
200
Time, min
Fig. 1. The mean venous blood concentration-time profile of galactose after quick intravenous injection of 0.5 g/kg galactose in 55 normal healthy volunteers and 36 patients with cirrhosis.
the slope, intercept and percent CV of each calibration curve was monitored. Six healthy subjects received galactose independently every morning which was re peated on consecutive days. The GSP. GEC and the modified GEC MGEC values were recorded and com pared as this activity occurred. GEC. GEC was calculated according to the follow ing formula, which was modified from Tygstrup’s [20] equation GEC = — —— (mg/kg-min). tc . o + 7 where D is the injected amount of galactose. tc.o is the amount of time required for the concentration to reach zero, extrapolated from a linear regression of the blood concentration-time curse from 20 min after injection to 60 min (usually, the galactose concentration at 60 min was higher than 2.22 mmol/1). and 7 is the empirical correction for uneven distribution in the body. MGEC. Galactose disposition in humans follows two pharmacokinetic open models: one rapid distribu tion phase followed by an elimination phase (fig. I ). Under the saturated dose of 0.5 g/kg. the linear portion (saturated portion) of galactose blood level in normal humans and in patients was different (fig. 1.2). Signifi cant variations in the distribution phase and the satu-
The Novel Liver Function Test GSP
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/5/2018 10:13:40 AM
Study Design All patients and volunteers underwent 3 min of quick intravenous administration of 0.5 g/kg galactose solution (0.4 g/nil). A butterfly needle was placed in a forearm vein. Volunteers and patients were kept in bed to rest for 3 h. The galactose tests consistently began at 9 a.m. and were completed by 12 a.m. The GEC was assessed largely according to Tygstrup’s [ 18] procedure [12. 16] using venous blood samples (0.5 ml) drawn at 5. 10. 15, 20, 25. 30. 35, 40, 50. 60. 70, 80, 90. 100, 110, 120, 130, 140. 150. 160, 170 and 180 min after the injection. Blood samples were kept in the ice bath until measured by an enzymatic method [17], A dupli cate. six-concentration calibration curve, ranging from 49.82 to 1.90! pg/ml, was constructed using blank blood drawn before administering to each subject. Within-day variation was evaluated by standard de viation and percent coefficient of variation (CV) for each concentration. A maximal of 10% CV was per mitted. Day-to-day variation was also checked by com paring the slope and intercept of the calibration curves.
Time, min
Time, min
3 26
5
10 20 2 5 3 0 40 5 0 00 70 SO 9 0 100
Time, min
2
rated portion among volunteers and patients were also noted (fig. 2). The rationale for the estimation of tc =o by extrapolating the regression line from 20 to 60 min was based on the assumption that the saturated por tion of the galactose-time curve was between 20 and 60 min. MGEC was estimated using each person’s linear portion of galactose concentration-time curve which is the real saturated portion after the distribution phase. GSP Method. After observing 52 normal healthy volunteers and 36 patients with hepatic cirrhosis in the MGEC study, highly significant different galactose blood levels were found between healthy people and patients at 40, 50 and 60 min following galactose administration (fig. 3). The galactose blood concentration was measured I h after administration and was selected as the GSP which measured the quantitative liver function.
Statistical Analysis Statistical evaluation utilized the analysis of vari ance (Anova) followed by the multiple range test, least significant difference (LSD), linear regression. Stu dent’s t test and Pearson's correlation coefficient test.
Results The sex. age and the possible etiology are presented in table 1. One normal subject, 2 with cirrhosis and 2 HCC patients did not complete the study. Results show that virus hepatitis is the major cause of liver dysfunc tion in Taiwan. The precision of the galactose
225
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/5/2018 10:13:40 AM
0
Fig. 2. The galactose saturation time period in nor mal volunteers (a) and patients with cirrhosis (b) obtained from each person’s galactose concentration time curve after receiving quick intravenous adminis tration of 0.5 g/kg galactose. Fig. 3. Difference of galactose blood concentration at different time points between 51 normal healthy volunteers and 36 patients with cirrhosis after 0.5 g/kg galactose was infused intravenously for 3 min. The concentrations 40. 50 and 60 min after infusion give highly significant differences between these two groups.
Table 1. Sex. age and/or the possible etiology of normal subjects and patients with CH. cirrhosis and HCC
Normal CH Cirrhosis HCC
Sex M
F
mean ± SD
range
HBV
HCV
alcoholic others
51 73 34 39
4 0 2 2
23.4 ±3.6 25.6 ±9.2 40.8 ±17.6 50.8 ±14.8
19-35 17-68 19-67 32-76
0 65 26 29
0 5 6 4
0 0 2 5
calibration curve was excellent with less than 6.66% within-day correlation coefficient (n = 3). The slope of each between-day calibration curve was compared, and the mean and stan dard derivations of slopes were recorded and monitored (fig. 4). Over 90% of the slopes fall in the range of the mean ± 2 SD. The mea sured and added galactose concentrations were also monitored for the accuracy of the assay. The estimated repetitive GSP. GEC and MGEC values for 6 healthy subjects who received galactose independently every morn ing on 2 consecutive days are listed in table 2. The CV for GEC. MGEC and GSP ranged from 0.64 to 16.87%. Table 3 shows the results of GSP, GEC and MGEC for normal healthy volunteers, pa tients with CH. cirrhosis and HCC. Highly significant differences were observed among these subjects (Anova, p < 0.001). Similar results were obtained using GEC and MGEC methods. Patients with HCC and cirrhosis were not significantly different (p > 0.05: using GSP: also there was no highly sig nificant difference (p < 0.05 only) using MGEC and GEC (table 3). Highly significant GSP values were observed among normal subjects, patients with CH, cirrhosis and HCC (p < 0.001). However, patients with CH and HCC showed no highly significant differ ences (p < 0.01 only) using GEC. This was
226
Possible etiology
Age, years
Tang/Hu
0 3 2 3
Table 2. The repeated estimated GSP, GEC and MGEC values on 2 consecutive days in 6 healthy sub jects
Subject
GSP pg/ml (% CV)
GEC mg/ml-kg (% CV)
MGEC mg/mbkg (% CV)
1
388-464 (8.92%) 111.4-156.6 (16.87%) 259.8-328.2 (11.63%) 129.1-178.9 (16.17%) 124.7-151.3 (9.64%) 228.74-241.26 (2.66%)
5.12-5.78 (6.06%) 6.97-7.41 (3.06%) 5.99-6.55 (4.42%) 6.97-7.37 (4.1%) 6.96-7.32 (2.52%) 6.21-6.29 (2.64%)
5.12-5.78 (5.97%) 7.47-8.97 (9.12%) 5.53-6.07 (4.62%) 7.25-7.73 (3.24%) 5.80-6.52 (5.38%) 6.03-6.35 (2.62%)
2 3 4 5 6
% CV was calculated using the standard deviation on 2 consecutive days of GSP. GEC or MGEC values divided by their respective means.
even worse while using MGEC. Nonsignifi cant results were obtained between patients with CH and HCC. GSP had highly significant correlations with both GEC (fig. 5) and MGEC (p < 0.001) with the correlation coefficients being 0.909 and 0.788. respectively: MGEC also
The Novel l iver Function Test GSP
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/5/2018 10:13:40 AM
Subjects
Fig. 4. Slope (mean ± SD and mean ± 2SD) of calibration curves from 41 subjects' own blank whole blood.
Fig. 5. The correlation between GEC and GSP among 173 healthy subjects and patients with CH. cir rhosis and HCC.
Table 3. Values and statistical results of GSP, GEC and MGEC (means ± SE) in 46 normal subjects (N). 66 patients with CH. 33 cirrhosis (C) and 36 HCC after 0.5 g/kg galactose was intravenously administered
Normal
CH
Cirr hosis
HCC
Statistical analysis: Anova & LSD N-CH
N-C
N-HCC
CH-C
CH-HCC C-HCC
GSP pg/ml-kg
247.2 ±16.5
423 ±26.0
630 ±41.0
580 ±40.9
< 0 .0 1
< 0.001
< 0.001
< 0.001
< 0.001
NS
GEC mg/ml •kg
6.68 ±0.07
5.80 ±0.11
4.78 ±0.17
5.28 ±0.18
< 0 .0 1
< 0.001
< 0.001
< 0.0 0 1
< 0 .0 1
< 0.05
MGEC mg/ml-kg
7.18 ±0.15
5.75 ±0.15
4.72 ±0.23
5.49 ±0.27
< 0 .0 1
