U270-9139/87/0703-0464$02.00/0 HEPATOLOGY Copyright cc; 1987 by the American Association for the Study of Liver Diseases

Val. 7. No. 3, pp. 464-467, 1987 Printed in U. S. A.

The Aminopyrine Breath Test Does Not Correlate with Histologic Disease Severity in Patients with Cholestasis

ALFRED L. BAKER, PATRICIA s. KRAGER, ALVIN N. KOTAKE AND DALE A. SCHOELLER Liver Study Unit and Section of Gastroenterology of the Department of Medicine, University of Chicago, Chicago, Illinois 60637

To determine whether the aminopyrine breath test can be used to document the presence of cirrhosis in patients with cholestatic liver disease, 19 patients (13 primary biliary cirrhosis, 4 sclerosing cholangitis and 2 chronic extrahepatic bile duct obstruction) underwent clinical and biochemical evaluations, liver biopsies and an aminopyrine breath test. Results were compared with those in 10 patients with biopsy-proven chronic active hepatitis with bridging and/or cirrhosis and in 22 healthy subjects. The aminopyrine breath test results in the 10 cholestatic patients with cirrhosis were not sig- nificantly different from the results in precirrhotic cho- lestatic patients (mean f S.D., 11.2 f 5.0 vs. 11.6 f 2.8 %dose per 2 hr, p > 0.05) or healthy subjects (1 1.5 2 2.9 %dose per 2 hr). In contrast, the results in the patients with chronic hepatitis were markedly depressed (3.2 f 1.9 %dose per 2 hr, p < 0.05). The aminopyrine breath test results did not correlate with results of conventional liver function tests in the cholestatic patients. These results demonstrate that the aminopyrine breath test is not clinically useful in identifying the presence of cir- rhosis in patients with cholestatic liver disease, and provide further evidence that decreased microsomal enzyme function is a late feature of cholestatic liver disease.

The clinical and biochemical features of alcoholic liver disease and chronic hepatitis generally cannot be used to determine the severity of these disorders, but the ami- nopyrine breath test (ABT) is markedly depressed in patients with histologic evidence of alcoholic cirrhosis or chronic active hepatitis with bridging or cirrhosis (1-3). The test is also a useful indicator of survival in alcoholic hepatitis and in patients with drug-induced hepatitis caused by acetaminophen overdose (1,4). The results of the ABT reflect hepatic microsomal enzyme function (5 , 6) and have also been useful for monitoring the effects of a number of drugs on hepatic drug metabolism. For example, oral contraceptives, disulfiram and cimetidine inhibit hepatic microsomal enzyme function, whereas phenobarbital, alcohol and rifampin enhance hepatic

Received January 21, 1986; accepted December 22, 1986. This work was supported in part by National Institutes of Health

Research Grant AM26678 and The Liver Research Fund, Chicago, Illinois.

Address reprint requests to: Alfred L. Baker, M.D., University of Chicago, Box 400, 5841 South Maryland Avenue, Chicago, Illinois 60637.

microsomal enzyme function, as reflected by changes in the ABT (7-10).

Although few patients with cholestatic liver disease have been studied systematically with the ABT, present evidence suggests that the effects of cholestasis on he- patic microsomal enzyme function are more variable than are chronic hepatitis and alcoholic cirrhosis (1 1- 13). The purpose of the present study was to determine whether the ABT could reliably identify the presence of cirrhosis in patients with cholestatic liver disease.

MATERIALS AND METHODS Nineteen patients (17 females and 2

males), aged 23 to 68 years, were studied. All patients had cholestatic liver disease (13 primary biliary cirrhosis, 4 scleros- ing cholangitis and 2 chronic extrahepatic bile duct obstruction) documented by clinical features and biochemical testing, as well as by liver biopsy performed percutaneously or a t the time of exploratory laparotomy. The six patients with sclerosing cholangitis and chronic extrahepatic bile duct obstruction had typical cholangiographic features of these diseases. Three pa- tients had ascites, each of whom had histologic evidence of cirrhosis on liver biopsy, but the other 16 patients did not have clinical features to suggest the presence of advanced liver disease. None were taking drugs which are known to affect hepatic microsomal function.

Liver function tests (serum bilirubin, alkaline phosphatase, AST, ALT, serum albumin, globulin and prothrombin time) and mitochondria1 antibody measurements were performed by the hospital clinical laboratory on each patient at the time of breath testing. Liver biopsy tissue was stained with hematox- ylin-eosin, Masson's trichrome, Perl's iron and reticulin stains. For the purpose of this study, biopsies were interpreted under code by one of us (A.L.B.) as demonstrating cirrhosis when portal-to-portal fibrosis was present throughout the biopsy along with at least one regenerative nodule.

Fifteen patients had breath tests at the time of liver biopsy, and four patients were tested within 3 months of exploratory laparotomy when dietary intake had returned to base level. The ABT was performed between 800 and 1000 in the morning with the patient resting in bed, according to a protocol approved by the University of Chicago Clinical Inves- tigation Committee. Written informed consent was obtained for each test. [''CC]Aminopyrine (2 mg per kg body weight) was dissolved in 50 ml water and administered orally. Breath sam- ples were collected a t baseline and at 30-min intervals for 2 hr and analyzed as previously described (14). The per cent of label recovered per millimole of expired CO, was calculated from the quantity of labeled CO, per millimole of total CO, and the dose. The per cent dose excreted per hour was determined for each

Patients Studied.

ABT.

464

Vol. 7, No. 3, 1987 ABT A N D HISTOLOGIC DISEASE IN CHOLESTATIC PATIENTS

2 0 0 -

10 0 -

I6 0 -

- r 140- cu al

al u) 0

L

Q 1 2 0 -

10.0- w C al 2 al 0 ° - a

m 6 0 -

Y

t- U

4 0 -

2 0-

465

sample, assuming a CO, production of 300 mM per m2 body surface area. Cumulative excretion over 2 hr was calculated by the trapezoid method, and the results were expressed as per cent dose excreted for 2 hr.

Ten patients with liver biopsy-proven chronic active hepa- titis with bridgmg and/or cirrhosis, determined by standard clinical and histologic criteria (151, also had an ABT following written informed consent and served as a comparison popula- tion. Only four had clinical evidence of cirrhosis (ascites and esophageal varices). Twenty-two healthy subjects also had an ABT and provided a control population, as reported elsewhere (16). These individuals had no history or clinical evidence of liver disease, and none were taking drugs during the study. Results of the ABTs in the healthy subjects were log-trans- formed for calculation of the mean + S.D. to minimize the influence of outlying values (16).

Comparisons between means were made using the Student’s t test for ungrouped data or the two- way analysis of variance for continuous variables or the xz test for categorical variables. A value of p < 0.05 was considered significant.

RESULTS Liver Morphology, Biochemical Data and ABT Results

in Patients with Chokstatic Liver Disease and Chronic Active Hepatitis. The results of the liver chemistry tests performed in this study are shown in Table 1. As ex- pected, the mean serum bilirubin levels were somewhat higher, the prothrombin times somewhat more prolonged and the serum albumin levels slightly depressed in the cholestatic cirrhotic patients compared to the cholestatic precirrhotic patients; none of these values were signifi- cantly different among the three groups studied, however (Table 1). Likewise, the occurrence of mitochondria1 antibodies and the alkaline phosphatase values were similar in the two groups of patients with cholestasis (Table 1). Thus, these tests could not be used to identify the presence of cirrhosis in these patients with choles- tasis (Table 1).

The results of the ABT were not significantly different in the cirrhotic cholestatic patients when compared with the precirrhotic cholestatic patterns (Figure 1). Fifteen of the 19 cholestatic patients had ABT results which fell between 6.5 and 19.1 %dose per 2 hr (mean -+ 2 S.D. of log-transformed results in the healthy control subjects) (Figure 1). Two patients, one with cirrhosis due to pri-

& U t i S t i d A d y s i s .

0

0

6. 0

0

0

0 8

! L)

T r I

Healthy Patients With Patients With Subjects Cholestasis Chronic Hepatitis

FIG. 1. ABT results in cholestatic patients compared to values obtained in healthy subjects and patients with chronic active hepatitis with bridgmg or cirrhosis. Closed circles represent values from patients with cirrhosis. Open circles represent values from precirrhotic patients. The horizontal line represents the mean, and the dashed lines represent ~2 S.D. of log-transformed results from healthy subjects. Two values from noncirrhotic cholestatic patients fell below 6.5 %dose per 2 hr, and two values fell about 19.1 %dose per 2 hr.

mary biliary cirrhosis and a precirrhotic patient with sclerosing cholangitis, had values of 20.8 and 19.2 %dose per 2 hr, respectively, and two precirrhotic patients with primary biliary cirrhosis had a value less than 6.5 %dose per 2 hr. In contrast, all but one of the patients with chronic active hepatitis with bridging or cirrhosis had values less than 6.5 %dose per 2 hr; only a single patient with bridging but no clear-cut evidence of cirrhosis had a value which fell within the range of normal (Figure 1). There was no correlation between the results of the ABT

TABLE 1. Liver chemistry tests in patients studied (mean f S.D.) Chronic active

Cholestasis. Cholestasis. hepatitis with precirrhotic cirrhotic bridges andlor p value’

cirrhosis

No. of patients Serum bilirubin (0.1-1.2 mg/dl) Alkaline phosphatase (10-85 IU) AST (<35 IU) ALT ( 4 5 IU) Serum albumin (3.2-5.0 gm/dl) Serum globulin (1.1-2.4 gm/dl) Prothrombin time (sec prolonged) Mitochondria1 antibody (no. positive)

9 3.02 f 1.81 301 f 202 102 f 42 120 f 50

3.60 f 0.41 4.0 2 0.78

1.07 f 1.33 5

10 5.36 f 7.70 529 f 400 116 f 42 132 f 48

3.58 f 0.77 3.94 f 0.33 1.49 ? 1.89

”

10 2.2 f 1.3 115 f 36 148 f 86 199 2 113

3.04 f 0.90 3.86 f 0.71 2.10 f 2.0

0

N S b c0.05

NS NS NS NS NS

C0.05

a The null hypothesis was that the results of each test were distributed evenly among the three groups. NS = not statistically significant.

466 BAKER ET AL. HEPATOLOGY

24.0- 52 - ?

!? 20.0-

- c a 2 16.0-

.-

.- - - i5

5 12.0-

a, a Q 0.0-

0 I-

L - - c -

4.0 - -

and the standard liver function tests in either the cir- rhotic or precirrhotic patients (Figure 2).

DISCUSSION The results of the present investigation suggests that

the ABT is not a useful indicator of the presence of cirrhosis in patients with cholestatic liver disease. Ten of the patients in the present study had histologic evi- dence of cirrhosis, yet none of the ABTs in these patients fell below the discriminatory level of the results in the 22 healthy subjects. In addition, the ABT results were not significantly different in the cirrhotic and precir- rhotic cholestatic patients. These investigations extend the previous observations of Burnstein and Galambos (111, who evaluated the ABT in 10 patients with primary biliary cirrhosis; however, only two of these patients had histological evidence of cirrhosis, but none of the results in their patients fell below the range of normal. In contrast, Galizzi et al. (13) studied 21 primary biliary cirrhosis patients with the ABT, and the results in eight patients fell below the values obtained in their healthy subjects. Six of these eight patients were said to have histologic and biochemical evidence of late primary bili- ary cirrhosis. Previous studies have shown that, in pa- tients with primary biliary cirrhosis, successive total serum bilirubin levels greater than 10 mg per dl obtained 6 months or more apart identify patients who have a mean survival time of 17 months (range = 13 to 22 months) (17), but only two of the patients in our study had values above this level. The difference between the

v

a, cn Q Q c cn 0

CI

n

a 800-

a, C 3 600- A

.- - a

0 2 O. E 400-

a, 0 0 0 a 0 200 -

O d 00 O. 0.

0 0 0 8 om

1 1 1 1 I I ~ I I I I I I I I 1 I I I I

28.0 1 1

0

0

n

2

results in the present study and those reported by Galizzi et al. may be due, at least in part, to the presence of patients with more advanced disease in the previous study; unfortunately, these workers did not report details of histologic findings and biochemical testing in their patients with primary biliary cirrhosis. Taken together, these results suggest that hepatic microsomal enzyme function is preserved until the late stages of cholestatic liver disease, even when cirrhosis is present.

Two of the cholestatic patients in the present study, one cirrhotic and one precirrhotic, had ABT results greater than 2 standard deviations above the mean for the healthy subjects, similar to findings in two of the patients reported by Burnstein and Galambos (11). Nei- ther of our two patients with elevated ABT results were taking any drugs known to alter hepatic microsomal enzyme function. These results provide further evidence that cholestasis, even in the presence of cirrhosis, has an uncertain effect on hepatic microsomal function, and suggest the possibility that unknown endogenous and exogenous substances may be retained by the liver in some cholestatic patients and result in enhanced hepatic microsomal enzyme function. It is also possible that alterations in other pharmacokinetic factors might ac- count for the results, but we know of no evidence to support such a possibility.

While the present studies demonstrate that the ABT is not clinically useful for the identification of cirrhosis in patients with cholestatic liver disease, further inves- tigation of these patients with the ABT might still be

i4001

1000 12001 0

0

0 0

0

0

ABT (percent dose per 2h) ABT (percent dose per 2h) FIG. 2. (Left) Serum bilirubin as a function of the results of the ABT in cirrhotic (closed circles) and precirrhotic (open circles) patients.

(Right) Serum alkaline phosphatase is a function of the results of the ABT in cirrhotic and precirrhotic patients. The total serum bilirubin and alkaline phosphatase showed no significant correlation with the results of the ABT (r = -0.384, p = not statistically significant; r = -0.491, p = not statistically significant). The additional liver function tests likewise failed to show a significant correlation with the results of the ABT (data not shown): AST (r = -0.148); ALT (r = 0.154); albumin (r = 0.091), and globulin (r = 0.109).

Vol. 7, No. 3, 1987 ABI' AND HISTOLOGIC DISEASE IN CHOLESTATIC PATIENTS 467

warranted. First, in patients with more advanced choles- tatic liver disease, the ABT might provide information on prognosis additional to the total serum bilirubin level. Such patients might be considered for new therapeutic agents or for hepatic transplantation. Second, the ABT might be useful in monitoring response to therapy in patients undergoing treatment designed to improve liver function, particularly in primary biliary cirrhosis where disease progression varies widely and therapeutic end- points are difficult to achieve. Finally, investigations of the mechanism of the elevated ABT results observed in some cholestatic patients might lead to further under- standing of human hepatic drug metabolism.

REFERENCES 1. Schneider JF, Baker AL, Haines NW, et al. Aminopyrine N-

demethylation: a prognostic test of liver function in patients with alcoholic liver disease. Gastroenterology 1980; 79:1145-1150.

2. Saunders JB, Lewis KO, Paton A. Early diagnosis of alcoholic cirrhosis by the aminopyrine breath test. Gastroenterology 1980;

3. Monroe PS, Baker AL, Schneider JF, et al. The aminopyrine breath test and serum bile acids reflect histologic severity in chronic hepatitis. Hepatology 1982; 2317-322.

4. Saunders JB, Wright N, Lewis KO. Predicting outcome of para- cetamol poisoning by using "C-aminopyrine breath test. Br Med J

5. Bircher J , Kupfer A, Gikalov K, et al. Aminopyrine demethylation measured by breath analysis in cirrhosis. Clin Pharmacol Ther

7 9 112-1 14.

1980, 1~279-280.

1976 20484492.

6. Hepner GW, Vesell ES. Aminopyrine disposition: studies on breath, saliva, and urine on normal subject.. and patients with liver disease. Clin Pharmacol Ther 1976; 20654-660.

7. Hepner GW. Vesell ES. Assessment of aminopyrine metabolism in man by breath analysis after oral administration of "C-aminopy- rine. Effects of phenobarbital, dilsulfiram. and portal cirrhosis. N Engl J Med 1974; 291:1384-1388.

8. Piken E, Hepner GW. Decreased hepatic microsomal reserve in patients with cirrhosis. Studies using aminopyrine as model drug. d Lab Clin Med 1979; 94:947-954.

9. H e n R, Koelz HR, Haemmerli UP, et al. Inhibition of hepatic demethylation of aminopyrine by oral contraceptive steroids in humans. Eur J Clin Invest 1978 8:27-30.

10. Desmond PV, Patwardhan R, Parker R, et al. Effect of cimetidine and other antihistamines on the elimination of aminopyrine, phen- acetin, and caffeine. Life Sci 1980; 26:1261-1268.

11. Burnstein AV, Galarnbos dT. ["CIAminopyrine breath test in chronic liver disease. Dig Dis Sci 1981; 261078-1083.

12. Hepner GW. Vessel1 ES. Aminopyrine metabolism in the presence of hyperbilirubinemia due to cholestasis or hepatocellular disease. Combined use of laboratory tests to study disease-induced altera- tions in drug disposition. Clin Pharmacol Ther 1977; 21:620-626.

1.3. Galizzi J, Long RG, Billing BH, et al. Assessment of the ("(3- aminopyrine breath test in liver disease. Gut 1978; 1940-45.

14. Schoeller DA, Klein PD. A microprocessor controlled mass spec- trometer for fully automated purification and isotopic analysis of breath CO,. Biomed Mass Spectrum 1979; 6350-355.

15. De (;root J, Desrnet VJ, Gedigk, et al. Acute and chronic hepatitis revisited. Lancet 1977; 2:914-919.

16. Schoeller DA, Baker AL, Monroe PS, et al. Comparison of different methods of expressing results of the aminopyrine breath test. Hepatology 1982; 2:455-462.

17. Shapiro dM, Smith H, Schaffner F. Serum bilirubin: a prognostic factor in primary biliary cirrhosis. Gut 1979 20137-140.