A. Norman

Elimination of theobromine metabolites in healthy adults

The metabolism of theobromine (TB) (500 mg per os) was determined by measuring plasma and saliva concentrations of TB and its metabolites 0-24 h after the load, and urinary excretion 0-48 h after the load. TB and its six metabolites were separated and quantified by combining high performance liquid chromatography and capillary electrophoresis. The urine analyses showed that unchanged TB accounted for 21 +/- 4% (mean +/- SD) of total excretion, the remainder being 7-methylxanthine (7-X, 36 +/- 5%), 3-methylxanthine (3-X, 21 +/- 4%), 6-amino-5[N-methylformylamino]-1-methyluracil (6-AMMU, 11 +/- 4%), 7-methyluric acid (7-U, 10 +/- 2%), 3,7-dimethyluric acid (3,7-U, 1.3 +/- 0.6%) and 3-methyluric acid (3-U, 0.5 +/- 0.4%). In addition to TB, 7-X and 3-X were consistently found in plasma and saliva; 6-AMMU and 7-U were found in plasma and saliva at concentrations < or = 2 mumol l-1 and 0.2 mumol l-1, respectively. TB concentrations in plasma and saliva were similar, whereas the saliva concentrations for 7-X and 3-X were found to be 63 +/- 17% of the plasma concentrations for 7-X and 74 +/- 13% for 3-X, respectively. The high urinary-to-plasma concentration ratio of 7-U (200-300) suggests high excretion of 7-U by the kidneys. Excretion of 7-X, 3-X and 6-AMMU was also high (urinary-to-plasma concentration ratio 45-150), whereas the excretion of TB was significantly lower than its metabolites (urinary-to-plasma concentration ratio 4-6). N3-demethylation of TB accounted for 58 +/- 7% of the urinary metabolites, N7-demethylation for 27 +/- 6%, C8-oxidation of 7-X for 22 +/- 4%, C8-oxidation of 3-X for 2 +/- 2% and formation of 6-AMMU for 13 +/- 4%. The ratio of N3- to N7-demethylation of TB remained constant during the load, but the large interindividual variation observed in ratio makes it unsuitable as a function test for evaluation of liver disease.

Assessment of dimethylxanthine formation from caffeine in healthy adults: comparison between plasma and saliva concentrations and urinary excretion of metabolites.

Caffeine (CA), paraxanthine (PX), theobromine (TB) and theophylline (TP) were determined in plasma (i.e. total concentrations), ultrafiltrate of plasma (free concentrations) and saliva, by isocratic high performance liquid chromatography (HPLC), 0-24 h after a 200 mg oral load of CA in 10 healthy adults. Total metabolism of CA was established by determination of urinary metabolites, 24 h after CA ingestion, by gradient HPLC and capillary electrophoresis. Saliva concentrations of CA, PX, and TP were lower than plasma concentrations (p < 0.001), whereas TB concentrations in plasma and saliva were similar. Saliva concentrations of CA, PX, TB and TP were higher than the free plasma concentrations (p < 0.001). The area under the plasma concentration-time curve (AUC) showed that PX accounted for 63 +/- 13% of the dimethylxanthines in plasma, while TB accounted for 27 +/- 15% and TP for 10 +/- 2.6%. In contrast, the urine analyses showed that 78 +/- 11% of the excreted metabolites were metabolized through the PX pathway, 14 +/- 8% through the TB pathway and 9 +/- 4% through the TP pathway. The percentage of the AUC for PX, TB and TP in plasma was similar to the percentage of each dimethylxanthine excreted unmetabolized in urine. The percentages of the AUC for PX and TB were correlated to the percentages of PX (r = 0.78, p < 0.001) and TB (r = 0.88, p < 0.001) in urine. The AUC for PX in plasma was lower than (p < 0.001) and correlated to the total PX pathway value (r = 0.95, p < 0.001) and the value for PX plus its specific metabolites in urine. The AUC for TB in plasma was higher than (p < 0.001) and correlated to the total TB pathway value (r = 0.93, p < 0.001) and the value for TB plus its specific metabolites in urine. The AUC for TP in plasma was similar to both the TP pathway value and the value for TP plus its specific metabolites in urine. It is concluded that the AUC for dimethylxanthines in plasma underestimates the formation of PX, overestimates the formation of TB and gives a similar formation of TP from CA, as judged from the urinary metabolites formed through the PX, TB, and TP pathways.

The nature of urinary bile acid conjugates in patients with extrahepatic cholestasis.

Urinary bile acids from patients with extrahepatic cholestasis were extracted with Sep-pak C18 cartridges and group separated on diethylaminohydroxypropyl Sephadex LH-20. The nature of the different conjugates of cholic and chenodeoxycholic acid in the fractions was studied after further separation by preparative thin-layer chromatography. The free and glycine-conjugated bile acids were quantified by capillary gas chromatography and identified by gas chromatography-mass spectrometry (GC/MS). Taurine conjugates were split with cholylglycine hydrolase and the liberated free bile acids analysed by GC/MS. Sulphate esters were hydrolysed with Helix pomatia and the resulting bile acid derivatives were analysed as above. After hydrolysis with cholylglycine hydrolase, the glucuronides of the unconjugated bile acids were separated and identified by GC/MS. Amino acid analysis of the different fractions revealed that glycine and taurine were the only amino acids present in connection with cholic and chenodeoxycholic acid. Large amounts of monosulphated bile acid conjugates were present but no disulphates. Only 3-sulphates were found. Both sulphates and glucuronides were found exclusively as glycine or taurine conjugates and no such derivatives of unconjugated bile acids were isolated. The isolated conjugates were split either by a combination of acid solvolysis and alkaline hydrolysis or by Helix pomatia and cholylglycine hydrolase.

Serum Concentrations and Excretion of Bile Acids in Cirrhosis

Bile acid concentrations in serum, and urinary and faecal excretion of bile acids have been studied in ten patients with liver cirrhosis as a consequence of alcohol abuse. Eight of the patients were categorized as Child group A, whereas the remaining two patients comprised Child group C. Individual bile acids were isolated and identified by gas chromatography coupled to mass spectrometry. Total fasting serum bile acid concentrations were elevated in all patients, but not correlated to conventional tests of liver function. Eight of the patients had increased urinary excretion of bile acids. Faecal bile acid-excretion was highly variable between patients, and also between Childs group A and C patients. Total fasting serum bile acid concentrations were not correlated to either urinary, faecal, or total bile acid excretion (= synthesis of bile acids) or to the ratio between urinary and faecal excretion of bile acids. The daily synthesis of bile acids showed a large overlap between Childs group A and C patients. The percentage of chenodeoxycholic acid and its metabolites relative to total daily excretion of bile acids did not correlate, indicating that the synthesis pathways for the primary bile acids does not systematically change in relation to the rate of synthesis. We conclude that even in mild cirrhosis, serum bile acid concentrations are elevated. However, no consistent changes in synthesis of bile acids or synthesis pathways was observed in such patients.

Caffeine metabolism in patients with chronic liver disease

An oral load of 200 mg (1030 mumol) caffeine (CA) was given to 13 patients with chronic liver diseases and to 11 healthy controls. The metabolism of CA was determined by following plasma concentrations and urinary excretion of CA and its metabolites. In addition, [2-14C]-caffeine was given orally to six patients to confirm the excretion through the different pathways. CA and its 14 main metabolites were separated and quantified by high performance liquid chromatography and capillary electrophoresis. Median (interquartile range) half-lives of CA were 19 (6.3-32) h in the patients and 3.8 (3.4-4.7) h in the controls. The wide range in the patients indicated varying degrees of liver dysfunction. Only 3 (2-4)% of administered CA was excreted unmetabolized in urine in the controls and the main degradation was through the paraxanthine (PX) pathway 82 (75-83)%. The combined theobromine (TB) and theophylline (TP) pathways (TB + TP) accounted for 15 (13-21)% of CA metabolism. Although the excretion of unmetabolized CA in the patients 6 (3-8)%, was similar to that in the controls, the metabolism through the PX pathway, 62 (44-65)%, decreased (p < 0.01 vs. controls), whereas the metabolism through the TB + TP pathways increased to 33 (30-47)%, p < 0.01. In controls, N3-, N7- and N1-demethylations were observed in 86 (83-89)%, 66 (62-70)% and 13 (9-18)%, respectively, of excreted metabolites. In patients the N3-demethylations, 71 (66-77)%, and N7-demethylations, 54 (48-59)%, decreased (p < 0.01 vs. controls), whereas N1-demethylation increased 30 (21-46)%, p < 0.01. The major C8-oxidation reaction, the oxidation of 1-methylxanthine, increased in patients (p < 0.01). We conclude that the slowed metabolism of CA in chronic liver disease is due to reduced N3- and N7-demethylations affecting biotransformation through the PX pathway.

Determination of caffeine and its metabolites in urine by high-performance liquid chromatography and capillary electrophoresis

Caffeine (CA) and its 14 main metabolites were determined in urine by reversed-phase high-performance liquid chromatography (RP-HPLC) and capillary electrophoresis (CE). After addition of 1,3,9-trimethylxanthine, uracil and beta-hydroxyethyltheophylline as internal standards, samples were separated by RP-HPLC into three fractions; A, B and C. The fractions were concentrated by lyophilization and analysed quantitatively by CE. Fraction A contained 5-acetylamino-6-amino-3-methyluracil, 3-methyluric acid, 7-methyluric acid and 1-methyluric acid. Fraction B contained 7-methylxanthine, 3-methylxanthine, 3,7-dimethyluric acid, 1-methylxanthine, 1,3-dimethyluric acid and 3,7-dimethylxanthine, while fraction C contained 1,7-dimethyluric acid, 1,7-dimethylxanthine, 1,3-dimethylxanthine, 1,3,7-trimethyluric acid and caffeine itself. The detection limit for the various metabolites ranged from 2-5 mumol l-1. The within-run precision for the metabolites ranged between 3.6% and 15.2%. The combination of HPLC and CE techniques was found to be a practical and specific method for determination of CA and its metabolites in urine.

Serum bilirubin subfractions in patients with alcohol abuse during detoxication

Sera were obtained from 41 alcohol abusers consecutively admitted for detoxication. Blood samples were withdrawn on the second, fourth and seventh days of abstention. Initial bilirubin values were moderately elevated in 10 patients. Determination of the bilirubin subfractions by high performance liquid chromatography showed elevated values of unconjugated (alpha), monoconjugated (beta), diconjugated (gamma) and albumin-bound (delta) bilirubin, in 8, 15, 4 and 15 patients, respectively. During abstention, the total bilirubin value normalized rapidly, only three patients still had values above the upper reference limit after 7 days. In patients with initially elevated values of bilirubin subfractions, only a few had elevated beta and gamma levels on the seventh day, whereas delta levels decreased at a slower rate and remained virtually unchanged. On admission, 27 patients exhibited elevated levels of serum bile acids; these values decreased during abstention and after 7 days only six patients had slightly elevated values. Only five patients had initially elevated levels of alkaline phosphatase (ALP). These became normalized in all but two patients during abstention. The results suggest that mild cholestasis is common among alcohol-abusers without clinically evident liver disease and that these changes are reversible on abstention.

Presence of bile acid metabolites in serum, urine, and faeces in cirrhosis

Studies have been made of the presence of bile acid metabolites in ten patients with liver cirrhosis as a consequence of alcohol abuse. Eight of the patients were categorized as Child group A, indicating only mild impairment of liver function, whereas the remaining two patients comprised Child group C. A complex mixture of bile acids was isolated from serum, urine, and faeces, and 26 bile acids were identified by gas-liquid chromatography coupled to mass spectrometry. Identification was made of the primary bile acids, cholic (C) and chenodeoxycholic (CDC) acid, and metabolites of these bile acids converted through 7-dehydroxylation, keto-formation, 6-hydroxylation, 1 beta-hydroxylation, allo-formation or nor-formation. All of the bile acids have previously been described either in healthy humans or patients with hepatobiliary disease. With the exception of C, CDC, and deoxycholic acid, all of the bile acids were present only infrequently, and none of the bile acids was pathognomonic for liver cirrhosis. The proportion of metabolites of the primary bile acids C and CDC was similar to that previously reported in healthy humans, the lowest proportion being recorded in the Child group C patients. Repeated determinations of the metabolite pattern in two patients showed large variations, indicating that the bile acid metabolism varies from time to time. We conclude that in mild cirrhosis, no significant alterations in microbial or hepatic transformation of bile acids seem to occur.

Lipoprotein-bound bile acids in serum from healthy men, postprandially and during fasting

Individual bile acids were determined by gas-liquid chromatography in, very low density, low density and high density lipoprotein fractions obtained by sequential ultracentrifugation of serum from normal adults, both postprandially and during fasting. The lipoproteins were found to contain 22-34% of fasting serum bile acids. The observed postprandial increase in bile acids did not exhibit any shift in the ratio between lipoprotein bound- and non-lipoprotein-bound bile acids. Bile acids were present in all isolated lipoprotein fractions, high density lipoproteins containing the highest amounts. In the lipoprotein fraction, a higher percentage of cholate than of chenodeoxycholate was found.

Bile constituents in ascitic fluid

Bile acids and other bile constituents were determined in serum and ascites from eight patients with liver cirrhosis and in ascites secondary to malignancy in six patients. In cirrhotic ascites, total bile acid levels averaged 53% of the serum levels. A positive correlation was evident between ascites and serum levels for both cholic and chenodeoxycholic acid. For cholic acid, the ascites to serum ratio was higher in all patients compared with the corresponding ratio for chenodeoxycholic acid. The ascites to serum ratios for glycine, taurine and sulphate conjugates were similar, no tendency being shown by any of the conjugates to leak more easily into ascites. The high levels of bile acids in cirrhotic ascites suggests that the abdominal cavity harbours a fraction of the bile acid pool, which should be taken into account when studying bile acid turnover in liver cirrhosis. Bilirubin levels in cirrhotic ascites averaged 24% of the serum values. A positive correlation between ascites and serum levels for unconjugated bilirubin was recorded, whereas the occurrence of bilirubin conjugates in ascites was variable. Albumin levels in cirrhotic ascites were 25% of the serum levels. The ascites to serum ratios for other proteins such as IgG, IgA and IgM and also cholesterol and phospholipids were lower than that for albumin. In malignant ascites, a pattern different from that in liver cirrhosis was seen, low bile acid levels being found. No difference between bilirubin levels was observed, while albumin and cholesterol levels were higher in malignant ascites, with no overlap between the patient groups. These results indicate that the complex mechanisms of ascites formation result in variable levels of bile constituents in ascitic fluid, which are further dependent on the underlying disease.