SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

Background:

Helicobacter pylori gastric infection has been associated with various digestive and extra-digestive diseases. The systemic influence of gastric H. pylori infection seems to be mediated by the release of various cytokines. In liver disease, bacterial infections have been associated with the impairment of liver metabolic function.

Aims:

To evaluate the influence of H. pylori infection on liver function as assessed by means of the monoethylglycinexylidide test, which depends upon liver blood flow and cytochrome P-450 activity, and the 13C-galactose breath test, which depends on cytosolic enzymatic activity and is correlated with hepatic functional mass. Moreover, to evaluate whether H. pylori-associated modifications of liver function may be related to tumour necrosis factor-α serum levels.

Patients and methods:

Thirty-five patients with liver cirrhosis of various aetiologies, who underwent monoethylglycinexylidide and 13C-galactose breath tests, were retrospectively evaluated for H. pylori infection by means of anti-H. pylori immunoglobulin G. The main clinical, biochemical and functional characteristics of the patients as well as their tumour necrosis factor-α serum levels were then analysed on the basis of the presence of H. pylori infection.

Results:

Twenty-one patients tested positive for H. pylori infection (60%), and 11 tested negative (31.4%). No clinical or biochemical differences were observed between H. pylori-infected and non-infected patients. H. pylori infection showed no difference in distribution according to Child–Pugh classes (A, 55%; B and C, 67%). The monoethylglycinexylidide test results were significantly lower at each sampling time in H. pylori-positive patients compared to H. pylori-negative patients (MEGX15, P=0.027; MEGX30, P=0.014; MEGX60, P=0.028), while 13C-galactose breath test showed no significant differences considering both cumulative percentage dose and percentage dose/h. The median tumour necrosis factor-α serum levels were no different between H. pylori-positive (16.1 pg/mL, 95% confidence interval, 8.7–28.7) and H. pylori-negative (12.3 pg/mL, 95% confidence interval, 8.7–23.4) patients.

Conclusions:

In cirrhotic patients, H. pylori infection seems to selectively affect cytochrome P-450 liver activity, while hepatic functional mass does not seem to be impaired. Tumour necrosis factor-α does not seem to be the mediator of this impairment. Further studies are needed to evaluate the impact of H. pylori eradication on parameters of liver function.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

Helicobacter pylori infection is implicated in the aetiology of chronic active gastritis,1 peptic ulcer disease, gastric adenocarcinoma and low-grade gastric B-cell lymphoma.2 Moreover, infection by H. pylori has been associated with other digestive and extra-digestive diseases.3 The mechanisms by which this bacterium apparently causes systemic illnesses probably include chronic inflammation and immune-mediated reactions caused by H. pylori infection in the gastric mucosa, leading to the production of several cytokines, such as tumour necrosis factor-α, interferon-γ and interleukins, responsible for the remote manifestations of disease.4, 5

Recent experimental studies have shown that infection by H. pylori and/or other Helicobacter spp. is associated with the development of chronic active hepatitis6 and with the production of liver-specific toxins.7 Moreover, it has been suggested that H. pylori vacuolating cytotoxin may reach the hepatocytes of patients suffering from both isolated hypertransaminasaemia and infection by cytotoxic strains of H. pylori.8

In cirrhotic patients, H. pylori infection has shown no correlation with the severity of liver disease, as assessed by the Child–Pugh score and by portal hypertension-related features, although contrasting results have been obtained with regard to hepatic encephalopathy.9–14 However, because the occurrence of systemic infections, as well as the increased permeability of the gut wall to bacterial antigens, in cirrhotic patients is associated with a decrease in liver function, probably mediated by pro-inflammatory cytokines,15–19 we deemed it of interest to evaluate whether H. pylori infection may influence liver function in patients with cirrhosis of the liver. We assessed liver function by means of the monoethylglycinexylidide (MEGX) test, which depends on liver blood flow and cytochrome P-450 activity,20–24 and by the 13C-galactose breath test (GBT), which depends on liver galactokinase activity25 (an enzyme located in the cytosol of hepatocytes)26 and the results of which are correlated with the hepatic functional mass.27 Colour-coded Doppler sonography was performed to evaluate the possible influence of liver blood flow on function tests. Moreover, we measured serum tumour necrosis factor-α levels to evaluate whether this cytokine may be involved in the relationship between H. pylori infection and liver function.

Therefore, we retrospectively analysed the H. pylori status in 35 cirrhotic patients of mixed aetiology, who had simultaneously undergone MEGX test, GBT and liver ultrasonography, so as to evaluate the influence of H. pylori gastric infection on various liver metabolic functions.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

The presence of immunoglobulin G (IgG) antibodies to H. pylori was retrospectively evaluated in 35 cirrhotic patients who were consecutively referred to our unit for the evaluation of liver disease. Hepatitis C virus infection was the cause of liver disease in 21 patients, hepatitis B virus infection in five, alcohol abuse in four, alcohol abuse and hepatitis C virus infection in two and cryptogenic cirrhosis in three. Patients with alcoholic liver disease had abstained from alcohol for at least 6 months prior to the study. The diagnosis of liver cirrhosis was made either histologically (n=2), or on the basis of pertinent clinical signs, Doppler ultrasonographic evidence of portal hypertension and/or the endoscopic presence of oesophageal or gastric varices. The clinical severity of liver disease was evaluated by means of the Child–Pugh score.28 All patients underwent upper digestive endoscopy, and none showed evidence of gastric or duodenal ulcer. None of the patients had previously undergone gastric surgery or therapy for H. pylori eradication, none was actively bleeding at the time of the study, and none was affected by diabetes mellitus. None of the patients was on drugs known to interfere with liver metabolic activity or had undergone antibiotic therapy within 6 months prior to the study.

The MEGX test, GBT and liver ultrasound examination with colour-coded Doppler sonography were simultaneously performed on all patients while resting and after an overnight fast. Blood samples for biochemical and anti-H. pylori IgG determination were obtained on the same day.

Biochemical, anti-H. pylori IgG and tumour necrosis factor-α assessment

Blood samples to evaluate aspartate aminotransferase, alanine aminotransferase, γ-glutamyl transpeptidase, alkaline phosphatase, bilirubin, albumin, γ-globulin, prothrombin activity, anti-H. pylori IgG and tumour necrosis factor-α were obtained after overnight fasting.

Routine biochemical tests were carried out using commercially available kits. Serum samples for anti-H. pylori IgG and tumour necrosis factor-α assessment were stored at –80 °C and analysed successively. A commercial enzyme-linked immunoabsorbent assay (GAP-IgG Helicobacter pylori, BIO-RAD, Milan, Italy) was used to detect serum anti-H. pylori IgG antibodies. The sensitivity, specificity and accuracy of the assay are 99.4%, 93.5% and 97.5%, respectively. All samples were analysed in duplicate and the results were expressed as the mean of the measurements. The coefficient of variation was 6.0%. According to the manufacturer’s instructions, serum anti-H. pylori IgG antibody values < 12.5 were considered to be negative, values between 12.5 and 20 equivocal, and values > 20 positive.

Blood samples for tumour necrosis factor-α determination were available in 27 patients. Serum tumour necrosis factor-α was measured by a commercial enzyme-linked immunoabsorbent assay kit (TNFα ELISA KIT, Diaclone Research, Besançon, France). All samples were analysed in duplicate according to the manufacturer’s instructions and the results were expressed as the mean of the measurements (pg/mL). The coefficient of variation was 15.0%.

MEGX test

All patients underwent MEGX test as a dynamic test to evaluate liver function.21, 23 MEGX formation from lidocaine depends upon liver blood flow and functioning mass.22 Lidocaine metabolization is carried out within the liver by the enzymatic system of cytochrome P-450.20, 24

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 15, 30 and 60 min after the end of the lidocaine infusion. The MEGX concentration at each sampling time was calculated as follows: MEGX0 – MEGXt. MEGX was measured using fluorescence polarization immunoassay with the TDX-system (Abbott Laboratories, North Chicago, IL, USA).

GBT

The GBT was performed as follows. Two basal breath samples were collected after an overnight fast. Then, 10 g/m2 body surface of 13C-galactose (1-13C-D-galactose, Euriso-Top Carbon 13 Breath Tests Substrates, Saint Aubin, France, supplied by Cortex, Milan, Italy), dissolved in 100 mL of sterile water, was orally administered to the patients. The concentration of labelled galactose was 1%. This concentration has been proven to ensure significant isotopic CO2 enrichment while limiting the costs related to the labelled substrate.29 The amount of galactose administered was chosen in order to saturate the galactose metabolic pathway so as to reflect the hepatic metabolic activity rather than the liver blood flow.27 Breath samples were collected every 30 min for 3 h after 13C-galactose administration and were obtained as follows. Patients blew through a small plastic tube for 10 s directly into a vial that was immediately sealed. The ratio of 13CO2 to 12CO2 was determined for each sample with an isotope ratio mass spectrometer and the excess 13CO2 was calculated by the increase in the isotope ratio. The δ value obtained was converted to the percentage 13C, and the results are expressed as a percentage of the administered dose of 13C recovered per hour (percentage dose/h) and as the cumulative percentage of administered dose of 13C recovered over time (cumulative percentage dose), as described previously.30

Liver ultrasonography and colour-coded Doppler sonography

In all patients, liver ultrasonography and colour-coded Doppler sonography were performed by an experienced operator (P.B.) at the same time as the MEGX test and GBT. Ultrasonography was carried out using AU 570 Asynchronous (Esaote Biomedica, Genoa, Italy). The liver supplying vessels were examined either from a subcostal position or via an intercostal access. A 3.5-MHz convex-array transducer with a Doppler frequency of 5.2 MHz was used for colour-coded Doppler sonography. For portal vein and hepatic artery imaging, a mid-range value (1000 MHz) was selected for the pulse repetition rate and then adjusted to suit the velocity that was detected. Colour-coded Doppler sonography was used to measure the mean velocity of the portal vein (cm/s), the hepatic artery resistance index and the hepatic artery pulsatility index.

Statistical analysis

Statistical analysis was carried out using the Mann–Whitney U-test to compare the means and the chi-squared test or Fisher’s exact test to compare proportions. Data are shown as the mean ± standard deviation (s.d.), except for tumour necrosis factor-α values which are expressed as the median and interquartile range. P < 0.05 for two-sided tests was considered to be statistically significant. Statistical analysis was performed using the MedCalc 5.00 statistical software (MedCalc Software, Mariakerke, Belgium).

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

Table 1 shows the main clinical and functional characteristics of the 35 cirrhotic patients. Twenty-nine patients were males (83%). Of these patients, 21 tested positive for anti-H. pylori IgG (60%), and 11 tested negative (31.4%). Three patients (8.6%) had equivocal anti-H. pylori IgG serum levels and were therefore not included further in the analysis.

Table 1.  Main characteristics of the 35 cirrhotic patients Thumbnail image of

The analysis of the patient characteristics was carried out by subdividing them according to positivity/negativity for H. pylori infection. Table 2 shows the clinical and biochemical characteristics of the patients thus subdivided. We observed no differences in gender distribution or age between H. pylori-infected and non-infected patients. None of the biochemical parameters evaluated was significantly different between H. pylori-positive or H. pylori-negative cirrhotic patients. H. pylori infection showed no difference in distribution on the basis of the Child–Pugh classes (A, 55%; B and C, 67%; chi-squared test, N.S.).

Table 2.  Characteristics of the patients subdivided on the basis of positivity for anti-H. pylori immunoglobulin G Thumbnail image of

The results of the patients’ liver function tests are shown in Table 3. The MEGX test results were significantly lower at each sampling time in H. pylori-positive patients compared to H. pylori-negative patients, while GBT did not show any significant difference between H. pylori-infected and non-infected patients considering both the cumulative percentage dose and percentage dose/h.

Table 3.  Results of monoethylglycinexylidide test and 13C-galactose breath test subdivided on the basis of positivity for anti-H. pylori immunoglobulin G Thumbnail image of

Serum tumour necrosis factor-α was measured in 27 patients (16 H. pylori-positive and 11 H. pylori-negative). Median serum tumour necrosis factor-α levels were no different between H. pylori-positive (16.1 pg/mL; interquartile range, 8.8–28.5) and H. pylori-negative (12.3 pg/mL; interquartile range, 10.6–20.1) patients.

Colour-coded Doppler sonography parameters (portal vein velocity, hepatic artery resistance index and hepatic artery pulsatility index) were not significantly different in the two subgroups of patients (H. pylori-positive vs. H. pylori-negative; Table 4).

Table 4.  Results of colour-coded Doppler sonography subdivided on the basis of positivity for anti-H. pylori immunoglobulin G Thumbnail image of

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

To date, studies conducted on cirrhotic patients with various aetiologies of liver disease have reported a prevalence of H. pylori infection ranging from 26% to 89%.14, 31 Furthermore, many of these studies took into consideration the influence of H. pylori infection on the presence of peptic ulcer disease, oesophageal varices and hypertensive gastropathy.9–11, 13, 14, 31, 32 Results have shown no evidence that H. pylori infection affects the presence and severity of liver disease-related events, although it seems that it might play a role in a subset of patients with hepatic encephalopathy.12, 33 Overall, these and other studies have shown no association between gastric H. pylori infection and the severity of liver disease as assessed by means of the Child–Pugh score.

Nevertheless, both human and experimental studies have shown various influences of H. pylori on the liver. Indeed, H. pylori can secrete cytotoxic factors which are capable of inducing cytopathic effects in liver cell lines,7 while, in humans, a mild increase in transaminases may be present during infection by cytotoxic strains of H. pylori.34 Moreover, H. pylori infection supposedly plays a role in patients with unknown causes of hypertransaminasaemia,8 and recent experimental studies have shown that infection by H. pylori and/or other Helicobacter spp. is associated with the development of chronic active hepatitis.6

Because several bacterial infections have been associated with the impairment of cytochrome P-450-dependent liver metabolism in liver disease,35, 36 and because exposure of the liver to prolonged inflammatory stimulation may cause metabolic hepatocellular dysfunction,36, 37 we deemed it of interest to evaluate whether chronic H. pylori infection may interfere with liver metabolic activity in cirrhotic patients. In order to do this, the MEGX test and GBT were chosen because they evaluate different liver functions. Indeed, while lidocaine metabolization to MEGX depends primarily on liver blood flow and cytochrome P-450 activity, galactose metabolism is carried out by galactokinase which is a cytosolic enzyme. Lastly, as tumour necrosis factor-α is implicated in both H. pylori infection-related extra-gastric diseases3–5, 38 and infection-associated liver dysfunction,16, 36 we measured its serum levels in order to evaluate its role in the modification of liver function.

Our study showed that the prevalence of H. pylori infection observed (60%) is in keeping with that previously seen in studies carried out on cirrhotic patients in Italy.11, 14 Moreover, we confirmed the lack of any relationship between H. pylori infection and the clinical severity of the disease as assessed by means of the Child–Pugh classification.

However, the main issue of our study is that H. pylori infection in cirrhotic patients seems to selectively affect liver cytochrome P-450 activity. Indeed, we showed that H. pylori-infected cirrhotic patients had lower mean MEGX values compared to non-infected patients, while GBT was not influenced by the presence of H. pylori infection. These findings suggest that, in cirrhotic patients, liver cytochrome P-450 activity may be depressed, while functional hepatic mass is not affected by H. pylori infection. Moreover, similar results in terms of sonographic Doppler parameters of liver supplying vessels in both H. pylori-positive and H. pylori-negative patients further rule out possible influences of liver blood flow on the MEGX test results.

These findings may be due to the fact that gastric H. pylori infection is associated with a local immuno-mediated inflammatory response.4, 5 This may be responsible for lesions which are remote from the primary site of infection, probably because of an increase in serum levels of various pro-inflammatory substances and cytokines.38, 39 In cirrhotic patients, hypothetically, chronic gastric H. pylori infection may cause a decrease in cytochrome P-450 activity by increasing the amount of cytokines and other mediators brought to the liver by the portal blood flow. Indeed, the impairment of cytochrome P-450-mediated metabolism during the course of infections has been attributed to the release of cytokines, such as tumour necrosis factor-α and interleukin-6.40, 41

In this study, we have shown that as far as tumour necrosis factor-α is concerned, this cytokine does not seem to play a major role in the alteration of cytochrome P-450 activity in H. pylori-positive cirrhotic patients. A possible explanation for this finding may be that serum tumour necrosis factor-α levels may not reflect the amount of this cytokine in the portal blood. Indeed, tumour necrosis factor-α produced as a consequence of the H. pylori-induced gastric inflammatory response may exert its action within the liver and could subsequently be inactivated.

Nevertheless, these mechanisms do not seem to be clinically relevant in normal subjects with H. pylori infection, whatever their origin. Indeed, we have previously shown that, in these subjects, the 13C-aminopyrine breath test, which shares common metabolic pathways with lidocaine, is normal.42

Although we have shown that, in H. pylori-infected cirrhotic patients, the hepatic functional mass does not seem to be affected by the presence of the bacterium, the finding that H. pylori infection is associated with the impairment of liver cytochrome P-450 activity may have clinical consequences. In fact, this enzyme is responsible for the metabolization of various substances.43 Thus, in cirrhotic patients, H. pylori infection may lead to further impairment of drug metabolism which is already present due to liver disease.

A possible drawback of our study is its retrospective setting. This did not allow us to evaluate liver function after eradication therapy. Indeed, the evaluation of liver function tests after eradication therapy would allow us to verify whether cytochrome P-450 activity could be improved by eliminating H. pylori. However, the probability of a lack of benefit of H. pylori eradication on liver function should be considered even in future studies.44

In conclusion, in cirrhotic patients, H. pylori infection seems to impair cytochrome P-450 liver activity, while hepatic functional mass does not seem to be significantly affected by this infection. In these patients, tumour necrosis factor-α does not seem to be responsible for this alteration. Further studies are needed to discover the possible mediators of H. pylori-associated impairment of cytochrome P-450 liver activity and to evaluate the impact of H. pylori eradication.

References

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. References