C-Galactose Breath Test and 13 C-Aminopyrine Breath Test for the Study of Liver Function in Chronic Liver Disease

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2005;3:279 285 C-Galactose Breath Test and C-Aminopyrine Breath Test for the Study of Liver Function in Chronic Liver Disease EDOARDO G. GIANNINI,* ALBERTO FASOLI,* PAOLO BORRO,* FEDERICA BOTTA,* FEDERICA MALFATTI,* ALESSANDRA FUMAGALLI,* EMANUELA TESTA,* SIMONE POLEGATO,* TIZIANA COTELLESSA,* SARA MILAZZO,* DOMENICO RISSO, and ROBERTO TESTA* *Gastroenterology Unit, Department of Internal Medicine, and Medical Statistics Unit, Department of Health Sciences, University of Genoa, Genoa, Italy Background & Aims: Liver biopsy examination is the gold standard to diagnose the presence of cirrhosis. The aim of this study was to evaluate the accuracy of both C-aminopyrine breath test ( C-ABT) and C-galactose breath test ( C-GBT) in the noninvasive assessment of the presence of cirrhosis in patients with chronic liver disease. Methods: We evaluated 61 patients with chronic liver disease of diverse etiologies (21 compensated cirrhosis). All patients underwent C-GBT and C- ABT, and the results were expressed as a percentage of the administered dose of C recovered per hour (%dose/h) and as the cumulative percentage of administered dose of C recovered over time (%dose cumulative). Results were analyzed according to absence vs presence of cirrhosis. Results: On average, C-GBT %dose/h and %dose cumulative were decreased significantly in patients with compensated cirrhosis, and the same finding was observed for C-ABT results from 30 to 120 minutes. C-GBT %dose/h at 120 minutes had 71.4% sensitivity, 85.0% specificity, and 83.7% accuracy, whereas C-ABT %dose cumulative at 30 minutes had 85.7% sensitivity, 67.5% specificity, and 77.1% accuracy for distinguishing between the 2 subgroups of patients. Combined assessment of C-GBT and C-ABT increased the diagnostic accuracy (80% positive predictive value) of either test alone and reached 92.5% specificity and 100% sensitivity for the diagnosis of cirrhosis. Conclusions: In patients with chronic liver disease, both C-GBT and C-ABT are useful for the diagnosis of cirrhosis. Combination of the tests increases the diagnostic yield of each test alone. The use of carbon breath tests ( C-BTs) for the noninvasive assessment of liver function in patients with chronic liver disease is not new, although recent studies have shown an increased interest in their clinical use. 1 6 This is likely the result of medical and financial changes that have occurred over time. First, the increase in the number of patients with hepatitis C virus (HCV) infection who are eligible for antiviral therapy has increased the need for a diagnostic aid that can help the physician to noninvasively stage and follow-up the disease. Second, improved efficacy of the antiviral therapy for HCV infection has suggested that liver biopsy examination may not always be needed in specific subgroups of patients before treatment. 7 Hence, in these patients, functional assessment that is correlated closely to histologic stage can be helpful to the clinician and is well accepted by the patients. Third, the diffusion of C- labeled substrates allows a higher number of patients to be tested without the risks related to the use of 14 C- labeled substrates. Furthermore, the availability of mass spectrometers for the diagnosis of Helicobacter pylori infection avoids the need of dedicated machines. Finally, the results of studies performed with various C compounds allows for the characterization of specific subfunctions of the liver (microsomes, cytosol, mitochondria) that can be altered differently over the course of liver disease or after liver transplantation. 1 6 In this changing environment, one of the greatest uses of C-BTs is the characterization of pathologic conditions that can be hard to distinguish on a clinical basis alone, eg, to identify the presence of well-compensated cirrhosis in patients with chronic liver disease, which often requires liver biopsy examination to solve the clinical question, 8 simple biochemical tests, sophisticated serum markers of fibrosis, and Doppler evaluation of liver blood-supplying vessels are inadequate. 9 C-BTs in this clinical situation are promising. 10 In this prospective study, we evaluated the usefulness and the accuracy of the C-galactose breath test ( C- GBT) and the C-aminopyrine breath test ( C-ABT) for the noninvasive identification of well-compensated cirrhosis in patients with chronic liver disease of diverse etiologies. Because C-GBT and C-ABT explore dif- Abbreviations used in this paper: C-BTs, carbon breath tests; C-ABT, C-aminopyrine breath test; C-GBT, C-galactose breath test; MEGX, monoethylglycinexylidide test; %dose/h, percentage of the administered dose of C recovered per hour; %dose cumulative, cumulative percentage of administered dose of C recovered over time. 2005 by the American Gastroenterological Association 1542-3565/05/$30.00 PII: 10.1053/S1542-3565(04)00720-7

280 GIANNINI ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 3, No. 3 ferent liver metabolic functions (cytosolic vs microsomal), whose activity can be altered in different ways in the course of liver disease, 11 we hypothesized that the results of the 2 assays could provide complementary rather than redundant information. Patients and Methods Patients Sixty-one patients were evaluated in this study. Among the study patients, 46 underwent liver biopsy examination to stage chronic alteration of liver enzymes, and 15 patients had previous histologic evidence of cirrhosis (diagnosed either during surgical procedures performed for other reasons or for staging of chronic alteration of liver enzymes). Patients with compensated cirrhosis but with previous episodes of decompensation (ascites, jaundice, hepatic encephalopathy, variceal bleeding), decompensated cirrhosis, diabetes, or ongoing alcohol abuse were not included in this study. The cause of liver disease was HCV in 45 patients, nonalcoholic steatohepatitis in 5 patients, hepatitis B virus in 4 patients, HCV and hepatitis B virus in 2 patients, primary biliary cirrhosis in 2 patients, autoimmune hepatitis in 2 patients, and alcoholic liver disease in 1 patient. None of the patients with HCV or hepatitis B virus was on, or had previously undergone, antiviral therapy. Patients with primary biliary cirrhosis were on ursodeoxycholic acid therapy, patients with autoimmune hepatitis were not on corticosteroid or immunosuppressive treatment, and the patient with alcohol-related liver disease had abstained from drinking alcohol for more than 6 months. None of the patients was on hepatotoxic drugs or drugs known to interfere with liver function. Some of the patients with HCV infection alone also took part in our previous study. 3 All patients underwent C-GBT, C-ABT, monoethylglycinexylidide (MEGX, a test of metabolic activity) test, and a comprehensive biochemical work-up that included assessment of platelet count, serum aspartate and alanine transaminase, serum albumin, serum bilirubin, and serum prothrombin activity. Moreover, all of the patients underwent abdominal ultrasonography including measurement of portal vein diameter (at the hepatic hylum) and spleen dimensions (bipolar diameter and surface area). Liver biopsy examination (n 46), C-GBT, C-ABT, biochemical assessment, and abdominal ultrasonography were performed within the same week. The C-GBT and MEGX test were performed contemporaneously, whereas C-ABT was performed 3 days apart to avoid possible influence on metabolic activity (MEGX) and CO 2 breath measurement ( C- GBT). 12 Informed consent was obtained from all patients before both liver biopsy examination and liver function tests. C-Galactose Breath Test The C-GBT was performed as follows: 2 basal breath samples were collected after an overnight fast, then 10 g/m 2 body surface of C-galactose (1- C D-galactose; Euriso-Top Carbon Breath Tests substrates, Saint Aubin, France, supplied by Cortex Italia, Milan, Italy) dissolved in 100 ml of sterile water were administered orally to the patients. The concentration of labeled galactose was 1%. This concentration has been proven to ensure significant isotopic CO 2 enrichment while limiting the costs related to the labeled substrate. The dose of galactose was chosen to saturate the galactose metabolic pathway so as to reflect the hepatic metabolic activity rather than liver blood flow, according to the method described by Shreeve et al. 14 Breath samples were collected every 30 minutes for 3 hours after C-galactose administration and were obtained as follows: patients blew through a small plastic tube for 10 seconds directly into a vial that was sealed immediately. C-Aminopyrine Breath Test All patients underwent C-ABT as follows: a basal breath sample was collected after an overnight (at least 12 hours) fast, then 2 mg/kg of C-aminopyrine (N-N-dimethyl- C-aminopyrine; Euriso-Top Carbon Breath Tests substrates, supplied by Cortex Italia) was dissolved in 200 ml of water and administered orally. Breath samples were collected every 30 minutes for 3 hours after C-aminopyrine administration and were obtained as follows: patients were asked to exhale for 10 seconds through a small plastic tube directly into a vial that was sealed immediately. C-Breath Test Assessment and Data Analysis The ratio of CO 2 to 12 CO 2 was determined for each sample with an isotope ratio mass spectrometer (Breath Mat; Finnigan, Bremen, Germany) and the excess CO 2 was calculated by the increase in the isotope ratio. The value that was obtained was converted to percent C and the results are expressed as a percentage of the administered dose of C recovered per hour (%dose/h), and in the cumulative percentage of administered dose of C recovered over time (%dose cumulative), as previously described. 15 Simplified formulas to calculate both %dose/h and %dose cumulative were used according to Ghoos. 16 The CO 2 production was estimated on the basis of body surface area assuming a CO 2 production of 5 mmol/m 2 /min. Patients were at rest for 15 minutes before the tests and remained at rest and fasted during the tests to minimize their total CO 2 production and to avoid the influence of food intake. Monoethylglycinexylidide Test All patients underwent the MEGX test contemporaneously to C-GBT. MEGX formation from lidocaine depends on liver blood flow and functioning mass. 17 Lidocaine metabolization is performed within the liver by the enzymatic system of cytochrome P-450. In this study, the MEGX test was used as an internal reference to assess the absence of flow dependence of C-GBT and confirm the correlation with C-ABT. 12 The MEGX test was performed as follows: lidocaine was administered at a dose of 1 mg/kg by slow (over a 2-min period) intravenous infusion. Blood samples were obtained 30 minutes after the end of the lidocaine infusion. The MEGX concentration at sampling time was calculated as follows: MEGX 30 MEGX 0. MEGX was

March 2005 C-BREATH TESTS AND CHRONIC LIVER DISEASE 281 measured using fluorescence polarization immunoassay with the TD X -system (Abbott Laboratories, North Chicago, IL). Statistical Analysis Statistical analysis was performed by using the nonparametric Mann Whitney U test for comparison of continuous variables and the 2 test for analysis of categoric variables. Receiver operating characteristics curves were used to identify the C-BTs cut-off levels with the best sensitivity, specificity, and accuracy for distinguishing the 2 subgroups of patients (without cirrhosis vs with cirrhosis). Correlation between 2 variables was assessed by using the Spearman rank correlation test (r s ). Demographic, biochemical, functional, and instrumental data of the patients are shown as mean and SD. Sensitivity, specificity, and accuracy are shown as absolute value and 95% confidence interval. A P value of.05 for 2-sided tests was considered statistically significant. Results Table 1 shows the main demographic features of the patients. As expected, cirrhotic patients were, on average, older than patients with chronic liver disease without cirrhosis. Sex distribution and anthropometric measurements of the patients were similar in the 2 subgroups of patients. On average, transaminase levels as well as parameters of synthetic liver function were not significantly different between patients with or without cirrhosis. Last and most noteworthy, mean portal vein diameter (11.3.23mmvs 11.1 2.8 mm) and spleen dimension (diameter, 11.5 1.7cmvs12.5 2.1 cm; surface area, 47.5 16.6 cm 2 vs 55.4 16.9 cm 2 ) were not significantly different in the 2 subgroups of patients. C-Galactose Breath Test The mean C-GBT %dose/h and %dose cumulative of the patients are shown in Figure 1. As can be Table 1. Main Clinical and Biochemical Features of the 61 Study Patients Subdivided According to the Presence of Cirrhosis Variable Chronic liver disease Without cirrhosis (n 40) With cirrhosis (n 21) Age (y) 48 11 56 11.02 Sex (men/women) 32/8 18/3 NS BMI (m 2 /kg) 25 3 24 2 NS AST (IU/mL) 72 55 98 75 NS ALT (IU/mL) 127 100 105 93 NS Bilirubin (mg/dl).8.3.8.3 NS Albumin (g/dl) 4.3.3 4.1.5 NS Prothrombin activity (%) 91 9 86 9 NS NOTE. Data are shown as mean and SD. Statistical analysis was performed by means of Mann Whitney U test or 2 test. BMI, body mass index; NS, not specified; AST, aspartate transaminase; ALT, alanine transaminase. P Figure 1. (A) C-GBT %dose/h and (B) %dose cumulative in patients with chronic liver disease without cirrhosis (solid line) and with compensated cirrhosis (dotted line). seen, a clear characterization of the profile and of the slope of the curves in the 2 subgroups of patients can be observed from the early sampling times. There was no correlation between age and C-GBT results when all the patients were studied as a whole or as separate subgroups. There was no correlation between C- GBT results and transaminase level, total bilirubin level, and prothrombin activity, whereas a trend toward a correlation was observed for serum albumin level (r s.218; P.09). There was no correlation at all between any C-GBT result and MEGX test result (r s range,.002.170). As shown in Table 2, both C-GBT %dose/h and %dose cumulative were significantly different in the 2 subgroups of patients at all sampling times. Because the C-GBT %dose/h and %dose cumulative at all sampling times were significantly different between patients without cirrhosis and those with Child s class A cirrhosis, we deemed it of interest to evaluate the sensitivity, specificity, and accuracy of C-GBT for the discrimination between the 2 subgroups of patients at the various sampling times. To perform this, we used C- GBT cut-off levels identified by means of receiver operating characteristic curves. This analysis showed that C-GBT %dose/h at a 120-minute cut-off level of 3.2

282 GIANNINI ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 3, No. 3 Table 2. C-GBT Results in the Groups of Patients With Chronic Liver Disease Without Cirrhosis and With Compensated Cirrhosis Chronic liver disease Time Without cirrhosis With cirrhosis P Cut-off level SS SP c-index %dose/h 30 2.2 1.3 1.0 1.0.001 1.2 76.2 80.0.804 60 4.6 2.6 2.3 1.7.001 1.8 61.9 87.5.804 90 6.0 3.1 2.9 2.4.001 2.0 61.9 95.0.816 120 7.1 3.8 3.2 2.7.001 3.2 71.4 85.0.837 150 7.5 4.3 3.1 2.2.001 7.3 100.0 52.5.833 180 6.7 3.9 3.0 2.8.001 2.2 57.1 92.5.801 %dose cumulative 30.3.2.1.1.001.1 66.7 77.5.758 60 1.1.6.6.4.001.4 57.1 85.0.777 90 2.4 1.3 1.3 1.0.001 1.1 66.7 85.0.785 120 4.0 2.1 2.1 1.6.001 1.4 57.1 95.0.798 150 5.8 3.1 3.0 2.3.001 2.0 57.1 95.0.801 180 7.5 4.0 3.8 3.0.001 2.4 52.4 97.5.799 NOTE. Data are shown as mean and SD. Statistical analysis was performed by means of Mann Whitney U test and receiver operating characteristics curves. SS, sensitivity; SP, specificity. had the highest accuracy (c-index,.837; 95% CI,.721.919) in differentiating the 2 subgroups of patients, with a sensitivity of 71.4% (95% CI, 47.8 88.6) and a specificity of 85% (95% CI, 70.2 94.3). The sampling time with the highest sensitivity was %dose/h at 150 minutes (cut-off level, 7.3; sensitivity, 100%; 95% CI, 94.1 100), whereas the sampling time with highest specificity was %dose cumulative at 180 minutes (cut-off level, 2.4; specificity, 97.5%; 95% CI, 86.8 99.6). All C-GBT %dose cumulative sampling times obtained an accuracy greater than 70%. In particular, the C-GBT %dose cumulative accuracy ranged from.758 (30 minutes, cutoff level,.1) to.801 (150 minutes, cut-off level, 2.0). C-Aminopyrine Breath Test Figure 2 shows the C-ABT %dose/h and %dose cumulative in the 2 subgroups of patients. As expected on the basis of our previous studies, the shape of the C-ABT curves of the 2 subgroups was similar, although cirrhotic patients showed decreased C-ABT results as compared with patients with chronic liver disease without cirrhosis. In particular, the C-ABT %dose/h values of cirrhotic patients were significantly lower at 30-, 60-, 90-, and 120-minute samples, whereas C-ABT %dose cumulative were significantly lower at all the sampling times (Table 3). Among the common biochemical parameters, C- ABT results correlated with aspartate transaminase level (eg, %dose/h at 30 minutes, r s.439, P.001), but not with alanine transaminase level, and showed a correlation with prothrombin activity (eg, %dose/h at 30 minutes, r s.398, P.002). The MEGX test results were correlated significantly with C-ABT results at every sampling time (highest correlation %dose/h at 60 Figure 2. (A) C-ABT %dose/h and (B) %dose cumulative in patients with chronic liver disease without cirrhosis (solid line) and with compensated cirrhosis (dotted line).

March 2005 C-BREATH TESTS AND CHRONIC LIVER DISEASE 283 Table 3. C-ABT Results in the Groups of Patients With Chronic Liver Disease Without Cirrhosis and With Compensated Cirrhosis Chronic liver disease Time Without cirrhosis With cirrhosis P Cut-off level SS SP c-index %dose/h 30 10.3 4.8 6.3 4.1.002 8.6 85.7 65.0.744 60 9.2 4.3 6.3 3.1.01 5.6 52.4 86.8.701 90 8.5 3.8 5.9 3.1.01 7.8 85.7 57.9.702 120 7.7 3.3 5.8 2.9.02 7.5 85.7 57.9.680 150 6.6 3.0 5.1 2.2 NS 180 5.9 2.9 4.4 2.4 NS %dose cumulative 30 1.5 1.0.8.5.001 1.1 85.7 67.5.771 60 3.9 1.8 2.4 1.4.001 3.2 85.7 63.2.754 90 6.1 2.8 3.9 2.1.002 5.3 85.7 60.5.743 120 8.1 3.7 5.4 2.8.004 7.1 85.7 60.5.730 150 9.9 4.6 6.7 3.3.006 8.6 81.0 60.5.718 180 11.5 5.3 7.9 3.8.008 11.1 85.7 52.6.709 NOTE. Data are shown as mean and SD. Statistical analysis was performed by means of Mann Whitney U test and receiver operating characteristic curves. Sensitivity, specificity, and accuracy (c-index) of sampling times that do not reach significant difference between the 2 subgroups are not shown. SS, sensitivity; SP, specificity; NS, not specified. minutes, r s.656, P.001), and although MEGX test results were significantly different in the 2 subgroups of patients (P.02), it had a lower diagnostic accuracy (c-index,.681) as compared with C-ABT. The C-ABT sampling time with the highest accuracy in discriminating between the 2 subgroups of patients was 30 minutes for both %dose/h (cut-off level, 8.6; c-index,.744; 95% CI,.616.847) and %dose cumulative (cut-off level, 1.1; c-index,.771; 95% CI,.645.869). All the C-ABT %dose/h and %dose cumulative that were significantly different between the 2 subgroups reached diagnostic accuracy greater than 68%. The sampling times with the highest sensitivity in differentiating the 2 subgroups were %dose/h and %dose cumulative at 30 minutes (85.7%; 95% CI, 63.6 96.8 for both measurements), whereas dose/h at 60 minutes had the highest specificity (86.8; 95% CI, 71.9 95.5). Combined Assessment of C-Galactose Breath Test and C-Aminopyrine Breath Test There was no correlation between C-GBT results and C-ABT results at each sampling time. Because C-GBT and C-ABT seemed to achieve complementary sensitivity and specificity for the diagnosis of compensated cirrhosis, we deemed it of interest to evaluate whether the combined assessment of the two C- BTs could increase the diagnostic accuracy. When taken into consideration alone, C-GBT %dose/h at 120 minutes (cut-off level, 3.2) and C-ABT %dose cumulative at 30 minutes (cut-off level, 1.1) were the results with the highest accuracy for diagnosis (Tables 2 and 3), and therefore were selected for combined assessment. Table 4 shows that the combined assessment of C- GBT and C-ABT actually increased the diagnostic accuracy for compensated cirrhosis. It is noteworthy that the presence of C-GBT %dose/h at 120 min greater than 3.2 and/or C-ABT %dose cumulative at 30 min greater than 1.1 virtually ruled out the presence of cirrhosis in 100% of the patients. Discussion The use of C-BTs with various substrates is not a novelty in clinical hepatology. However, the diagnostic aims should be focused clearly and the role of the tests should be well characterized when such important and sophisticated tools such as the C-BTs are proposed for clinical use. 17 20 Both recent and less recent human studies convincingly showed that C-BTs exploring either microsomal, cytosolic, or mitochondrial hepatocellular subfunctions are able to distinguish patients with various degrees of liver disease from normal subjects, as well as distinguishing patients with compensated cirrhosis from those with decompensated cirrhosis. 1 5,11 However, less convincing evidence has been provided for the identification of patients with well-compensated cirrhosis among patients with chronic liver disease. This differentiation is of particular concern for the clinician, and a liver biopsy examination often is required to solve this important clinical dilemma. As a matter of fact, wellcompensated cirrhosis can be difficult to diagnose on

284 GIANNINI ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 3, No. 3 Table 4. Diagnostic Accuracy of C-GBT and C-ABT, Alone or Combined, for Identifying the Presence of Cirrhosis in Patients With Chronic Liver Disease Sampling time Value (%) 95% Confidence interval Sensitivity C-ABT %dose cumulative 30 min 85.7 (63.7 97.0) C-GBT %dose/h 120 min 71.4 (47.8 88.7) C-ABT %dose cumulative 30 min or C-GBT %dose/h 120 min 100.0 (83.9 100.0) C-ABT %dose cumulative 30 min and C-GBT %dose/h 120 min 57.1 (34.0 78.2) Specificity C-ABT %dose cumulative 30 min 67.5 (50.9 81.4) C-GBT %dose/h 120 min 85.0 (70.2 94.3) C-ABT %dose cumulative 30 min or C-GBT %dose/h 120 min 60.0 (43.3 75.1) C-ABT %dose cumulative 30 min and C-GBT %dose/h 120 min 92.5 (79.6 98.4) PPV C-ABT %dose cumulative 30 min 58.1 (39.1 75.5) C-GBT %dose/h 120 min 71.4 (47.8 88.7) C-ABT %dose cumulative 30 min or C-GBT %dose/h 120 min 56.8 (39.5 72.9) C-ABT %dose cumulative 30 min and C-GBT %dose/h 120 min 80.0 (51.9 95.7) NPV C-ABT %dose cumulative 30 min 90.0 (73.5 97.9) C-GBT %dose/h 120 min 85.0 (70.2 94.3) C-ABT %dose cumulative 30 min or C-GBT %dose/h 120 min 100.0 (85.8 100.0) C-ABT %dose cumulative 30 min and C-GBT %dose/h 120 min 80.4 (66.1 90.6) NOTE. C-ABT %dose cumulative 30 min cut-off 1.1, C-GBT %dose/h 120 min cut-off 3.2. PPV, positive predictive value; NPV, negative predictive value. clinical grounds alone, even for the experienced clinician. 8,9 Clinicians are becoming reluctant to perform a liver biopsy when results do not modify the therapeutic approach, 21 and the procedure often is perceived negatively by the patient. 22 Recently, biochemical markers of liver fibrosis such as the Fibrotest (BioPredictive SAS, Paris, France) were proposed for the noninvasive staging of patients with liver disease. 23 Nevertheless, these markers seem to reflect liver fibrosis rather than actually assessing liver function. 9,20 We observed that both C-GBT %dose/h and C- GBT %dose cumulative were significantly different between patients with chronic liver disease without cirrhosis and Child s class A cirrhosis. C-GBT results at various sampling times after oral administration of C- Galactose were significantly lower in patients with compensated cirrhosis, and the discriminatory accuracy of C-GBT %dose/h in this analysis was greater than 80% at all sampling times. The fact that all C-GBT sampling times, regardless of the modality of expression of results, were significantly different between the 2 subgroups may enhance the reliability of the test results and suggests that the test can be shortened to 2 hours. Furthermore, we confirmed that the time course of C appearance in the breath after C-aminopyrine administration follows specific kinetics in the 2 subgroups of patients (without vs with cirrhosis), and that early C- ABT sampling times have important diagnostic accuracy, thus suggesting that in this setting they can be used without the risk of missing any information. Although the results of the 2 probes showed similar accuracy for diagnosing cirrhosis, we found no correlation among them. One interpretation is that the functions explored by the tests can be altered in different ways over the course of the disease, and that combined assessment of the two C-BTs could increase the diagnostic accuracy of the tests alone. Furthermore, there are other reasons supporting the use of multiple C- BTs. 18,20 First, the tests explore different liver functional activity (cytochrome P-450 and galactokinase) differently located within the hepatocyte (microsomes and cytosol). Second, the enzymatic functions that were explored may ( C-ABT) or may not ( C-GBT, under the circumstances of this study) be influenced by sex or xenobiotics. Actually, the combined assessment of C- GBT dose/h at 120 minutes and C-ABT dose/h at 30 minutes (the sampling times with the highest diagnostic accuracy) increased the discriminant ability for cirrhosis, and the 100% sensitivity allowed us to rule out the diagnosis in this population. This topic should be assessed in larger studies, and its cost-effectiveness should be evaluated. Furthermore, the use of specific cut-off levels needs to be validated in larger prospective studies before being proposed in clinical practice. The results obtained need to be examined with caution. We observed a certain degree of overlap among the C-BT results of the 2 subgroups of patients. This likely is owing to the fact that chronic liver disease is characterized by various degrees of fibrosis and necroinflammatory activity, and therefore the results obtained in patients with chronic

March 2005 C-BREATH TESTS AND CHRONIC LIVER DISEASE 285 liver disease and severe fibrosis can be no different from the ones obtained in some subjects with initial, well-compensated cirrhosis. C-BT results need to be interpreted from a clinical point of view. C-BTs can prevent the need for a liver biopsy in a substantial number of patients with chronic liver disease while they provide the clinician with comprehensive evaluation in patients who do not want to or cannot undergo a liver biopsy. The C-BTs may provide interesting, noninvasive insight into liver pathophysiology. 24,25 Approximately three quarters of the patients we studied had chronic viral hepatitis, and this could be considered a weakness of this study. However, chronic infection by hepatitis viruses is the leading cause of liver disease in many countries. 26 Further studies are needed to evaluate the generalizability of results to patients with disease causes other than viral hepatitis. Moreover, it is possible that in different clinical settings, other substrates might provide even better results. 6,11 In conclusion, our results showed that in patients with chronic liver disease of diverse etiologies, both C-GBT and C-ABT are able to identify a clinical condition such as compensated cirrhosis, which is not easy to discriminate on a clinical basis alone. 8 We found that cytosolic as well as microsomal subfunctions are altered in the relatively early phases of chronic liver disease, although our results seem to suggest that they can be deranged in different ways in the course of the disease. Increase in the diagnostic accuracy is achieved by the combined use of C-GBT and C-ABT, suggesting the 2 tests are complementary and not redundant. Our results provide a promising starting point for further studies. References 1. Mion F, Rousseau M, Scoazec J-Y, et al. C-Galactose breath test: correlation with liver fibrosis in chronic hepatitis C. Eur J Clin Invest 1999;29:624 629. 2. Lara Baruque S, Razquin M, Jimenez I, et al. C-phenylalanine and C-methacetin breath test to evaluate functional capacity of hepatocyte in chronic liver disease. Dig Liver Dis 2000;32:226 232. 3. Giannini E, Fasoli A, Chiarbonello B, et al. C-aminopyrine breath test to evaluate severity of disease in patients with chronic hepatitis C virus infection. Aliment Pharmacol Ther 2002;16:717 725. 4. Park GJ, Katelaris PH, Jones DB, et al. Validity of the C-caffeine breath test as a noninvasive, quantitative test of liver function. Hepatology 2003;38:1227 1236. 5. Ishii Y, Suzuki S, Kohno T, et al. L-[1- C] phenylalanine breath test reflects histological changes in the liver. J Surg Res 2003;114:120 125. 6. Petrolati A, Festi D, De Berardinis G, et al. C-methacetin breath test for monitoring hepatic function in cirrhotic patients before and after liver transplantation. Aliment Pharmacol Ther 2003;18: 785 790. 7. National Institute of Health. National Institute of Health Consensus Development Conference statement: management of hepatitis C: June 10-12, 2002. Hepatology 2002;36(Suppl 1): S3 S20. 8. Bain VG, Bonacini M, Govindarajan S, et al. A multicentre study of the usefulness of liver biopsy in hepatitis C. J Viral Hepat 2004;11:375 382. 9. Fontana RJ, Lok AS. Noninvasive monitoring of patients with chronic hepatitis C. Hepatology 2002;36(Suppl 1):S57 S64. 10. Mion F, Rousseau M. C-Galactose breath test: a sensitive test to measure liver function. Gut 1998;43(Suppl 3):S25 S26. 11. Armuzzi A, Candelli M, Zocco MA, et al. Breath testing for human liver function. Aliment Pharmacol Ther 2002;16:1977 1996. 12. Fasoli A, Giannini E, Botta F, et al. CO2 excretion in breath of normal subjects and cirrhotic patients after C-aminopyrine oral load. Comparison with MEGX test in functional differentiation between chronic hepatitis and liver cirrhosis. Hepatogastroenterology 2000;47:234 238.. Mion F, Géloën A, Rousseau M, et al. Mechanism of carbon tetrachloride autoprotection: an in vivo study based on C- aminopyrine and C-galactose breath tests. Life Sci 1994; 54:2093 2098. 14. Shreeve WW, Shoop JD, Ott DG, et al. Test for alcoholic cirrhosis by conversion of [14C] or [C] galactose to expired CO2. Gastroenterology 1976;72:98 101. 15. Schoeller D, Baker AL, Monroe PS, et al. Comparison of different methods of expressing results of the aminopyrine breath test. Hepatology 1982;4:455 462. 16. Ghoos Y. CO2-breath tests at the laboratory digestion-absorption. Belgium: University Hospital Gasthuisberg Leuven, 1996: 43 54. 17. Jalan R, Hayes PC. Review article: quantitative tests of liver function. Aliment Pharmacol Ther 1995;9:263 270. 18. Giannini E, Fasoli A, Botta F, et al. Is there a role for multiple quantitative liver function tests? Dig Liver Dis 2000;32:644. 19. Giannini EG, Testa R. Non invasive assessment of liver function: trying not to miss an opportunity. Aliment Pharmacol Ther 2004;19:827 828. 20. Giannini EG, Testa R. Breath tests and liver fibrosis. Eur Rev Med Pharmacol Sci 2004;8:51 54. 21. Andriulli A, Festa V, Leandro G, et al. Usefulness of a liver biopsy in the evaluation of patients with elevated ALT values and serological markers of hepatitis viral infection: an AIGO study. Dig Dis Sci 2001;46:1409 1415. 22. Andriulli A, Annese V, Facciorusso D, et al. First do no harm: power, oppression, and violence of liver biopsy. Gastroenterology 2003;125:272 273. 23. Imbert-Bismut F, Ratziu V, Pieroni L, et al, and the MULTIVIRC. Biochemical markers of liver fibrosis in patients with hepatitis C virus infection: a prospective study. Lancet 2001;357:1069 1075. 24. Scheuer PJ. Grading and staging of chronic hepatitis: a trap for the unwary. Available: http://www.gastrohep.com/freespeech/ freespeech.asp?id 24 16. Accessed: March 2001. 25. Garcia G, Keeffe EB. Liver biopsy in chronic hepatitis C: routine or selective. Am J Gastroenterol 2001;96:3053 3055. 26. Stroffolini T, Sagnelli E, Almasio P, et al, Italian Hospitals Collaborating Groups. Characteristics of liver cirrhosis in Italy: results from a multicenter national study. Dig Liver Dis 2004;36:56 60. Address requests for reprints to: Roberto Testa, MD, Division of Gastroenterology, Department of Internal Medicine, University of Genoa, Viale Benedetto XV, no. 6, 162, Genoa, Italy. e-mail: rtesta@unige.it; fax: (39) 010-353-7956. Supported by MURST 60% under grant from the Italian Ministry of Scientific and Technological Research. Presented in part at the 55th annual meeting of the American Association for the Study of Liver Diseases, October 29 November 2, 2004, Boston, Massachusetts.