Alcohol and Alcoholism Advance Access originally published online on July 11, 2008
Alcohol and Alcoholism 2008 43(5):559-563; doi:10.1093/alcalc/agn046
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Influence of Liver Biopsy on Abstinence in Alcohol-Dependent Patients
,1,2,1,4
1 Unités d’Hépatologie, Hôpital Cochin, Paris, France
2 Unité d’Addictologie, Hôpital Cochin, Paris, France
3 INSERM U567, Université René Descartes, Paris, France
* Corresponding author: Unité d’Hépatologie–Hôpital Cochin, 27, rue du Faubourg St Jacques, 75014 Paris, France. Tel.: +33-01-58-41-30-18; E-mail: stanislas.pol{at}cch.aphp.fr
Received 1 February 2008; first review notified 10 March 2008; in revised form 28 March 2008; accepted 3 April 2008
 |
ABSTRACT |
|---|
 TOP ABSTRACT IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Background: Liver biopsy indication for the evaluation of alcoholic liver disease is controversial. Our aim was to investigate the influence of the biopsy on the patients’ motivation for abstinence. Methods: We retrospectively analysed, in a population of 324 patients hospitalized for alcohol withdrawal, the impact of liver biopsy on the following clinical outcomes: rapid loss to follow-up (immediately after hospital discharge), early relapse (< 3 months) and long-lasting abstinence (> 12 months). The biopsy was performed in 136 patients who had liver enzymes perturbations. Hepatic lesions were graded as mild (isolated steatosis and/or non-bridging fibrosis), moderate (bridging fibrosis and/or moderate alcoholic hepatitis) or severe (cirrhosis and/or marked alcoholic hepatitis) in 66 (48%), 41 (30%) and 29 (21%) cases, respectively. Results: In univariate analysis, patients who had a liver biopsy were less likely to be rapidly lost to follow-up (12% versus 27%, P = 0.003) but had a lower rate of long-term abstinence (20% versus 34%, P = 0.025). In multivariate analysis, age was the only factor significantly associated with clinical outcome: older patients had higher rate of long-term abstinence (OR = 1.041; P = 0.010). Among patients who had a biopsy, those with severe hepatic lesions had a lower rate of rapid relapse than those with moderate or mild lesions (32% versus 68% and 56%, P = 0.018) but the rate of long-term abstinence was similar in the three groups. Conclusion: This observational study does not support the notion that liver biopsy has a significant influence on the maintenance of alcohol abstinence in patients with alcoholic liver disease.
|
Introduction |
|---|
|
TOPABSTRACTIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
In European countries, alcoholic liver disease (ALD) is the first aetiology of cirrhosis and alcoholic cirrhosis causes 40% of alcohol-related deaths (Expertise Collective INSERM, 2001
). Despite the importance of the problem, management of ALD remains challenging and controversial. One of the difficulties is the diagnosis of hepatic histological lesions that have important prognostic implications. Hepatic lesions of ALD range from isolated mild steatosis, which usually has a benign course over years, to cirrhosis and its risk of life-threatening complications such as liver cancer or hepatic failure. Massive steatosis, alcoholic hepatitis or extensive fibrosis can be seen as intermediary lesions with a significant risk of progression through cirrhosis (Mathurin et al., 2007). Those distinct histological lesions may be associated with similar clinical, radiological and laboratory tests findings. A liver biopsy is therefore often necessary to establish a precise histopathological diagnosis and some authors recommend its wide usage in patients affected by ALD (Morgan, 1999). This opinion is disputed by others, and a nationwide survey conducted in 1997 among 89 French hospitals has shown that despite its high prevalence among liver diseases, ALD represented only 10% of the 2084 liver biopsies performed (Cadranel et al., 2000). This may reflect the common opinion that biopsy results will not change the treatment that relies on alcohol abstinence whatever the severity of liver lesions. On the other hand, another common opinion is that the motivation of the patient for abstinence, the cornerstone of treatment success, may be influenced by a diagnosis of ALD established on precise histopathological criteria. There is very little data to evaluate the validity of this hypothesis. Even if several cohort studies have reported a significant rate of sustained abstinence in patients that had received a diagnosis of alcoholic cirrhosis (Gish et al., 2001; Iasi et al., 2003; Patek and Hermos, 1981), those studies never addressed specifically the role of the diagnosis on the rate of abstinence. In that context, our goal was to investigate the influence of a liver biopsy on abstinence in a population of patients hospitalized for alcohol withdrawal. |
Patients and Methods |
|---|
|
TOPABSTRACTIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
This was a retrospective study carried out in the Liver Unit of the Necker Hospital in Paris, France. From January 1995 to October 2001, 900 patients were hospitalized at least once for alcohol withdrawal. We randomly selected a third of those patients and excluded those who had already had a liver biopsy before the hospitalization. All the patients fulfilled the criteria for alcohol dependence according to the DSM-IV classification (American Psychiatric Association, 1994
). A liver biopsy was performed during the hospitalization in patients with liver enzyme perturbations. We retrospectively classified the liver biopsy results in three groups according to the severity of the cases. Normal histology, steatosis affecting <50% of hepatocytes and non-bridging fibrosis were classified as ‘mild lesions’. Minimal alcoholic hepatitis, steatosis affecting >50% of hepatocytes and fibrosis with few septa were classified as ‘moderate lesions’. Marked alcoholic hepatitis, fibrosis with numerous septa and cirrhosis were classified as ‘severe lesions’. Follow-up data were collected from the medical file or were pursued from the other institutions that took the patients in charge subsequently. Abstinence was defined as no subjective report or objective indication of alcohol consumption. Durations of follow-up and of abstinence were calculated from the day of hospital discharge. We studied the impact of liver biopsy on the following clinical outcomes: rapid loss to follow-up (immediately after hospital discharge), early relapse (<3 months) and long-term abstinence (>12 months). Long-term abstinence was chosen as the primary endpoint. Approval of the study was obtained from the local ethical board.
Quantitative parameters were described using mean and standard error while qualitative parameters were described as percentages. Transaminases (ASAT and ALAT) and gamma glutamyl transferase (
GT) values were expressed as multiples of the upper normal limit. Quantitative values were compared using the Student t-test and non-parametric tests (Mann–Whitney) when the data were skewed. Qualitative values were compared using the chi-square test or the Fisher exact test. Multivariate analysis was completed by using logistic binary regression. The statistical significance level was <5%. All analyses were done using the SAS software 8.0 (SAS Institute Inc., NC, USA) and the SPSS software 10.0 (SPSS Inc, Chicago, IL, USA).
|
Results |
|---|
|
TOPABSTRACTIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Characteristics of patients
Characteristics of patients are summarized in Table 1. Among the 327 randomly selected patients, 24 had already had a biopsy before the hospitalization and 303 were finally included in the study. Among those, 192 (58.7%) were male and the mean age at inclusion was 44 years. Sixty-seven percent of the patients had already been hospitalized at least once for alcohol withdrawal. Most of the patients (n = 261, 86%) were current smokers and 15 (5%) had an opioid substitution. An extra-hepatic complication of alcoholism (pancreatitis, neuropathy, epilepsy, optic neuritis or cerebellar degeneration) was present in 85 (28%) patients. Twenty-four patients (7.9%) had antibody against the hepatitis C virus (HCV), 19 (79%) of those having detectable virus in their blood by RT-PCR. Three patients (1.0%) were tested positive for HBs antigen, 39 (12%) were tested negative for HBsAg but tested positive for HBc antibodies and 45 (15%) had been vaccinated against the hepatitis B virus. Human immunodeficiency virus (HIV) infection was found in six patients (2.0%), two of those being co-infected with HCV.
|
During their hospitalization, one-third of the patients (n = 103, 34%) received intravenous fluid and vitamins. Benzodiazepines were prescribed to 273 patients (83%). Severe withdrawal symptoms (delirium or seizure) developed in 26 patients (8%), requiring a transfer in intensive care unit in 3 cases. Three patients (1.0%) received corticoids for severe alcoholic hepatitis with a favourable outcome in all cases. The average length of hospitalization was 11 ± 4 days. Nine patients (3.0%) left the unit without medical agreement or for disciplinary reasons. Three patients (1.0%) were transferred to a psychiatric unit.
Histology
A biopsy was performed in 137 patients (42%). In two patients (1.45%), the biopsy was complicated by a bleeding that required a transfer to an intensive care unit for observation. As detailed in Table 1, the patients who had a biopsy had significant higher daily alcohol intake (212 ± 102 versus 191 ± 108 g, P = 0.04), were more likely to experience withdrawal symptoms (46% versus 28%, P = 0.005), also had more extra-hepatic co-morbidity related to alcohol use (38% versus 20%, P = 0.0007), had more biological perturbations (P < 0.0001), were more often treated by intravenous infusion (48% versus 22%, P < 0.0001), had a longer duration of hospitalization (12.4 ± 4.3 versus 10.3 ± 4.3 days, P < 0.0001) and had more often undergone other invasive procedures (especially endoscopy: 79% versus 43%, P < 0.0001).
Microscopic examination revealed a normal liver in a single patient. Isolated steatosis was found in 31 patients (23%). Steatofibrosis was present in 68 patients (50%), acute alcoholic hepatitis in 27 (20%). Fifteen patients (11%) had cirrhosis. One biopsy (0.73%) has not yielded sufficient material for interpretation. Using the classification given in the ‘Patients and Methods’ section, we classified the liver lesions as mild, moderate and severe in 66 (48%), 41 (30%) and 29 (21%) cases, respectively. Characteristics of the patients according to the liver biopsy results are given in Table 2. Interestingly, there was no significant difference between severity of the liver lesions according to average daily alcohol intake, extra-hepatic complications of alcohol or withdrawal symptoms.
|
Rapid loss to follow-up
After hospital discharge, 61 patients (20%) were immediately lost to follow-up. Rapid loss to follow-up was significantly less frequent in the patients that had a liver biopsy (14.3% versus 31.7%, P = 0.009) or an upper gastrointestinal endoscopy (17% versus 37%, P = 0.005).
Abstinence
Among the 242 patients that were not immediately lost to follow-up, the average follow-up was 35 ± 27 months. In patients who had a biopsy, there was a trend for higher rate of early relapse (53% versus 44%, NS) and a significantly lower rate of long-term abstinence (20% versus 34%, P = 0.025) (Fig. 1A). As detailed in Table 3, other factors associated with long-term abstinence in univariate analysis were higher age (46 ± 8 versus 44 ± 9 years, P = 0.03), lower average daily alcohol intake (176 ± 105 versus 210 ± 136, P = 0.025), lower GT levels (7.3 ± 9.7 versus 4.4 ± 5.8 x ULN, P = 0.003) and the absence of extra-hepatic somatic complication of alcoholism (33% versus 17%, P = 0.017). In multivariate analysis, age was the only factor to remain significantly associated with long-term abstinence (OR = 1.041; P = 0.01).
|
|
If rapid loss to follow-up was assimilated to an early relapse (Fig. 1B), the rate of early relapse appeared to be identical in groups with or without biopsy (59%). The rate of long-term abstinence remained higher in the patients that did not have a biopsy (25% versus 18%) but that difference was no more significant.
Among patients who had a liver biopsy (Fig. 1C), those with severe hepatic lesions had a significantly lower rate of rapid relapse than those with moderate or mild lesions (32% versus 68% and 56%, P = 0.018). Severe hepatic lesions were also associated with a non-significant trend for a higher rate of long-term abstinence (29% versus 18% and 18%, NS).
Results were not modified when patients with ongoing chronic viral infection (HIV, HBV or HCV) were excluded from the analysis (data not shown).
Other clinical events
New hospitalization was necessary for 170 patients (56%), mostly for a new attempt of alcohol withdrawal. Three patients developed refractory ascites. One of those patients eventually achieved to maintain a prolonged abstinence with resolution of ascites. Seven patients died during the follow-up process. Three deaths were caused by end-stage liver failure and one by suicide.
|
Discussion |
|---|
|
TOPABSTRACTIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
This study addressed the performance of liver biopsy as a therapeutic intervention in patients suffering from ALD. From this strict point of view, liver biopsy appears to have, at best, a modest impact. According to our primary endpoint (long-term abstinence), biopsy was associated with a slightly worse outcome: among the 137 patients who had a biopsy, 24 (17%) were still abstinent after 1 year of follow-up against 41 of the 167 patients (25%) who did not have a biopsy. Multivariate analyses did not show a significant effect of biopsy on abstinence. Secondary endpoints were more favourable to liver biopsy: biopsy was associated with a lower rate of immediate loss to follow (17% versus 37%) and patients who had severe liver lesions had a lower rate of early relapse compared to those with moderate or mild lesions (32% versus 68% and 56%, respectively). The rate of long-term abstinence however remained low in the group of patients with severe lesions (less than one-third). This clearly shows that a diagnosis of a life threatening disease such as cirrhosis is often insufficient to prevent a relapse in alcohol-dependent patients. This finding is not a surprise in this population as continuation of a substance use ‘despite knowledge of having a persistent [...] physical [...] problem that is likely to have been caused [...] by the substance’ is one of the criteria for dependence in the DSM IV classification (American Psychiatric Association, 1994
). The impact of a liver biopsy may be different in a population of alcohol abusers without criteria for dependence.
It is also important to take into account that liver disease is, in many ways, difficult to understand for the patients who sometimes do not know the function of the liver or even its position in the body. Most of them do not know histopathological terms such as ‘steatosis’, ‘inflammation’" or ‘fibrosis’ and their clinical consequences. Furthermore, ALD remains asymptomatic during most of its natural history thereby contributing to render this pathology abstract to patients. Those points are of importance as chronic alcohol consumption is associated with neuropsychological deficits and difficulty of acquiring novel semantic knowledge (Pitel et al., 2007). Those learning disabilities may impair the assimilation of the results of the biopsy and their potential implications. Efforts to optimize the comprehension and the recall of this information by the patient could therefore improve the impact of a liver biopsy in the context of alcohol dependence. From this perspective, we can expect that educational, counselling and communication models developed for chronic illness (asthma, diabetes, etc.) could be used for alcohol-related disorders. In particular, alcoholic patients need to improve their knowledge of dependence and its consequences.
Our study has limitations. First, it was an observational study and multivariate analyses may have not controlled for all the potential bias. In particular, patients who had a biopsy had a higher alcohol average daily intake and a higher rate of physical symptoms of alcohol withdrawal. Those features probably indicate a more severe dependence to alcohol, and they may therefore have contributed to the higher rate of relapse in the biopsy group. Second, the analysis of the rate of relapse is probably an imperfect way to measure the influence of an intervention on patient motivation for alcohol abstinence. According to the model proposed by Prochaska and DiClemente (1986), the process of change can be seen as an evolution through a series of stages from pre-contemplation to maintenance. In the context of dependence, this evolution is rarely linear and most people do not manage to maintain their gain at first attempt. Relapses occur frequently and can be seen as an integral part of the process that is not necessarily equivalent to a treatment failure. In other words, even if biopsy realization and its results did not decrease the rate of relapse after the hospitalization, we cannot exclude that they had an influence on the patient that could play a role later in the patient history. From that perspective, biopsy-beneficial effect on the rate of loss to follow-up is interesting as assiduity is probably related to positive outcome in alcohol-dependent patients.
Enhancing the patient motivation for abstinence is not the only reason for performing a liver biopsy. First, in an alcohol-dependent patient with liver enzyme perturbations, a biopsy can participate to the exclusion of hepatic co-morbidity such as iron overload, chronic hepatitis, biliary disease, etc. Previous studies have shown that as much as 20% of patients having a history of alcohol abuse also have histopathological evidences of non-alcoholic liver disease (Levin et al., 1979). In our study, seven (5.8%) patients showed histopathological signs of non-alcoholic liver disease. In five patients, all known to have contracted hepatitis C, there were features of chronic hepatitis. A 39-year-old female patient had a peliosis of undetermined aetiology and a 46-year-old male patient had a massive iron overload on Perls coloration that led to the diagnosis of genetic haemochromatosis. Taken together, the interest of a liver biopsy in the aetiological diagnosis of liver disease in our alcohol-dependent patients appeared marginal compared to historical cohort. This is probably related to the progresses that have been made in the non-invasive diagnostic tools in the last decades. For example, now that intrahepatic iron can be quantified by magnetic resonance imaging and that molecular analyses can disclose mutations in the HFE gene, liver biopsy is no more a prerequisite for the diagnosis of genetic haemochromatosis. The most important role of a liver biopsy is probably the determination of the stage of fibrosis. A diagnosis of advanced fibrosis (as it was made in 11% of the patients who had a biopsy in our cohort) has major implications on the care of patients with ALD because of the risk of liver cancer that justifies periodic screening. Here also, the development of non-invasive methods for the diagnosis of fibrosis may diminish biopsy indications in the future. It is important to note however that most of those non-invasive methods have been validated in patients suffering from hepatitis C and that transposition to ALD must be made very cautiously.
To conclude, our study did not show a clear impact of liver biopsy on maintaining abstinence in alcohol-dependent patients. As liver biopsy is an invasive procedure, its indication must be carefully weighed in those patients. On the other hand, our results should not be viewed as an argument to neglect the somatic care of alcohol-dependent patients. First, the engagement in a process towards alcohol abstinence must remain a free decision of an informed individual. This is the role of the physician in charge of the patient to make sure that he is fully informed of the consequences of his alcohol consumption and a complete somatic checkup comes within this framework. Second, optimal somatic cares have their own justification that must not be conditioned to a complete abstinence that is of course essential but can come later in the history of the patient. Suboptimal care in the interval can have irreversible physical consequences that may hinder the patients to gain the full benefit of abstinence. Finally, we think that further studies are needed to understand and optimize the motivational impact of a somatic diagnosis in an alcohol-dependent patient.
|
FOOTNOTES |
|---|
Both authors contributed equally to this work. 
4 Present address: Pôle Addictologie Prévention Education, Centre Hospitalier Victor Jousselin, Dreux, France.
|
References |
|---|
|
TOPABSTRACTIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders (1994) Washington, DC: American Psychiatric Association.
Cadranel JF, Rufat P, Degos F. Practices of liver biopsy in France: results of a prospective nationwide survey. For the Group of Epidemiology of the French Association for the Study of the Liver (AFEF). Hepatology (2000) 32:477–81.[CrossRef][Web of Science][Medline]
Expertise Collective INSERM. Maladies alcooliques du foie. In: Alcool: Effets sur la santé (2001) (INSERM):35–65.
Gish RG, Lee A, Brooks L, et al. Long-term follow-up of patients diagnosed with alcohol dependence or alcohol abuse who were evaluated for liver transplantation. Liver Transpl (2001) 7:581–7.[CrossRef][Web of Science][Medline]
Iasi MS, Vieira A, Anez CI, et al. Recurrence of alcohol ingestion in liver transplantation candidates. Transplant Proc (2003) 35:1123–4.[CrossRef][Web of Science][Medline]
Levin DM, Baker AL, Riddell RH, et al. Nonalcoholic liver disease. Overlooked causes of liver injury in patients with heavy alcohol consumption. Am J Med (1979) 66:429–34.[CrossRef][Web of Science][Medline]
Mathurin P, Beuzin F, Louvet A, et al. Fibrosis progression occurs in a subgroup of heavy drinkers with typical histological features. Aliment Pharmacol Ther (2007) 25:1047–54.[Web of Science][Medline]
Morgan M. Alcoholic liver disease: natural history, diagnosis, clinical features, evaluation, management, prognosis, and prevention. In: Oxford Textbook of Clinical Hepatology—Bircher J, Benhamou J, McIntyre N, Rizzeto M, Rodés J., eds. (1999) 2. Oxford: Oxford University Press. 1183–238.
Patek AJ Jr, Hermos JA. Recovery from alcoholism in cirrhotic patients: a study of 45 cases. Am J Med (1981) 70:783–5.[Medline]
Pitel AL, Witkowski T, Vabret F, et al. Effect of episodic and working memory impairments on semantic and cognitive procedural learning at alcohol treatment entry. Alcohol Clin Exp Res (2007) 31:238–48.[CrossRef][Web of Science][Medline]
Prochaska J, DiClemente C. Towards a comprehensive model of change. In: Treating Addictive Behaviours: Processes of Change—Miller W, Heather N, eds. (1986) New York: Plenum.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||