| |
Drinking Alcohol and Risk for Progression of Liver Disease
|
| |
| |
SUMMARY BY JULES LEVIN: The study found for individuals drinking >50 grams of alcohol a day (5 drinks) there is an increased risk for progression of liver disease. Drinking >8 drinks per day increased risk even greater. However, the study did not find that drinking alcohol at levels less than that showed a statistically significant increased risk regardless of the level of drinking. In other words, if you drink 2 drinks per day the risk is not greater than if you don't drink at all or compared to 3 drinks per day. However, they saw a trend that increased drinking levels may increase the risk, so drinking 3 drinks per day may be worse than 1 drink. Below is an editorial on the study followed by a report from the study itself. Also below is a discussion of past studies including one in IDUs finding a risk for progression in lite drinkers.
Note from Jules Levin: For HCV-infected individuals with a history of substance or alcohol abuse, complete abstinence from alcohol is recommended.
"You Can Lead a Patient to Alcohol, But You Can't Make Him Fibrose"
April 2004, Hepatology, Editorial by Harvey Alter, CDC, Viral Hepatitis Editor, Hepatology
Comments by Monto et al (study authors)
"...a prospective study of American inner-city injection drug users, 33% of whom were coinfected with HIV, found a step-wise, statistically significant increase in the adjusted incidence of cirrhosis when >13 g/day and >37 g/day were consumed...a second, a retrospective study of Swedish patients, found increased, light alcohol consumption (median 5.7 g/day) in the 44 patients whose fibrosis worsened over time compared to the 34 patients (median 2.6 g/day) whose fibrosis did not worsen. Thus, the literature to date addressing light and moderate alcohol intake in chronic HCV is conflicting, but different endpoints and patient groups make interpretation of the data difficult.
EDITORIAL: The study of Monto et al. (see article report below) gives a sobering account of the effects of various amounts of alcohol intake on hepatitis C-related fibrosis. At issue is the compounding effect of alcohol and HCV and the practical concern of how much alcohol an HCV-infected patient can safely consume. The literature is replete with studies that uniformly show that high alcohol intake (>50-80 g/day) greatly exacerbates fibrosis progression and cirrhosis development in patients with HCV infection. The net outcome has always been much more than additive. Few studies, however, have systematically examined the effects of lesser quantities of alcohol, and in those that have, the results have been conflicting. Monto et al. performed a careful estimation of lifetime alcohol intake and derived an average consumption in grams per day based on the calculation that one drink equals 10 grams of pure ethanol. The cohort was then divided into light (0-20 g/day), moderate (20-50 g/day) and heavy (>50 g/day) consumption. The mean alcohol consumption in the cohort was 41.6 g/day, but it was not normally distributed so that the median was 17.4 g/d; 29% consumed >50 g/day. Interestingly, alcohol intake did not correlate with the ALT or the inflammatory score on liver biopsy, suggesting that alcohol does not exert its fibrotic effects by enhancing inflammation. Alcohol intake was associated with a stepwise increase in mean fibrosis, and if alcohol use was dichotomized into <50g/day versus >50 g/day, there was a significant difference (P = .01) between the groups. If dichotomized at >=80 g/day, the difference was even greater (P = .006), but no significant difference in fibrosis was found if dichotomization was performed at lower levels of intake. In a multivariate analysis, histologic inflammation, serum ALT, and patient age were independent predictors of fibrosis, but alcohol was not, though it was close (P = .06). Importantly, at every alcohol level, there was a broad range of fibrosis outcomes, and even among heavy drinkers, 47% had stage 0 or stage 1 fibrosis. Just as in persons not infected with HCV, there
appears to be a range of susceptibility to the fibrotic effects of alcohol.
Because there was a stepwise increase in the odds ratio for fibrosis with each level of alcohol intake, the authors conclude that even light alcohol intake may be playing a role in fibrosis development and that their study size of 800 may be inadequate to demonstrate a subtle effect of low amounts of alcohol on fibrosis. I think this is a very cautious interpretation of their data. One could also reasonably conclude that low amounts of alcohol are probably not detrimental to HCV-infected patients and that one drink a day may have more cardiac benefit than it has fibrogenic detriment, as consistent with the NIH consensus statement. I'll drink to that.
Risks of a range of alcohol intake on hepatitis C-related fibrosis
Hepatology
Volume 39, Issue 3, March 2004
Alexander Monto *, Keyur Patel, Alan Bostrom, Stephen Pianko, Paul Pockros, John G. McHutchison, Teresa L. WrightDepartment of Gastroenterology, University of California at San Francisco, San Francisco, CA
Significant alcohol use or alcohol use in the top 25% of patients (>59 g/day) compared to the lowest 25% (<4 g/day) was associated with increased fibrosis. There was not a significant association between alcohol and mean fibrosis (P = .08) or mean fibrosis progression (P = .62) overall. Also, each level of alcohol intake was not associated with significantly increased mean fibrosis compared to the next higher level (i.e., 0 vs. 0.1-20 g/day, 0.1-20 g/day vs. 20.1-50 g/day). Figure 1 shows the incremental effects of alcohol on fibrosis. Although overall fibrosis was greater in patients with HCV who drink heavily than in those who did not, there was a range of disease in each category of alcohol intake.
We have quantified alcohol intake in 800 patients undergoing liver biopsy at three sites. A history of heavy alcohol intake correlated with a higher degree of histological fibrosis. Light or moderate drinkers did not have statistically greater fibrosis than nondrinkers, but alcohol may play a role in their liver disease. At each level of alcohol intake, there is a broad range of liver disease; successively heavier intake leads to subtle increases in risk for fibrosis. Patient age, serum ALT, and histological inflammation retained independent correlations with fibrosis in multivariate analysis, whereas heavy alcohol intake (>59 g/day) did not. The variability in natural history for a given level of alcohol intake points to a central role for immunological and/or genetic variables in HCV disease progression.
For men, drinking 0.1 to 20 grams of alcohol per day (1 drink=10 grams) odds ratio for fibrosis compared to non-drinkers was 1.06; drinking 20 to 50 grams perday odds ratio increased to 1.19; drinking 50 to 80 grams increased odds ratio to 1.26; drinking >80 grams/day increased odds ratio to 1.76. Increases were not statisyically significant except for drinkers of >80 grams/day compared to non-drinkers. For women there did not appear to be such a stepwise increase in risk: for 0.1 to 20 grams/day risk was 0.79; 20 to 50 grams/day risk was 0.86; 50 to 80 risk was 1.39; but for >80 grams risk was 0.68.
Ethnicity, risk factor for HCV acquisition, and HCV genotype did not correlate with cirrhosis in univariate analysis.
One important question has been: is there a safe level of alcohol intake in patients with chronic HCV infection? This study does not find this to be the case. We did not find a statistically significant association between alcohol intake and mean fibrosis on liver biopsy until a consumption level of 50 g/day of alcohol, and this only in univariate analysis. In the cohort overall, however, both mean fibrosis and the odds ratio for fibrosis increased step-wise even among patients with less than 50 g/day of alcohol consumption.
Thus, light and moderate alcohol intake may be playing a role in fibrosis, but even with 800 patients, our cohort size may be inadequate to demonstrate the subtle effect of low amounts of alcohol on fibrosis. A safe level of alcohol intake is not demonstrated. Light and moderate intake exert less of an effect on fibrosis than heavy intake, however, and may indeed have minimal or no effect. Balancing this small risk of liver disease progression against potential cardiovascular benefit may be particularly pertinent to middle-aged men, who worldwide constitute the majority of patients with HCV, and who are also at high risk for cardiovascular disease. Risk-benefit assessment should be individualized for each patient.
Another finding of our study is that patients with hepatitis C may have differing susceptibility to the effects of alcohol. This is accepted in alcoholic liver disease in the absence of hepatitis C, where only a subset of heavy drinkers develop cirrhosis. In our study, patients with HCV who drink heavily have, on average, more liver disease than those who do not. Individual heavy drinkers, however, may have minimal liver disease, and in fact 47% of 'heavy drinkers' (>50 g/day) in this study had Stage 0 or Stage 1 liver disease. Potential explanations for this somewhat surprising finding include: differential susceptibility to alcohol; different patterns of alcohol intake (e.g., binge-drinking vs. habitual drinking), which were not captured fully by our questionnaire; reversal of adverse effects of alcohol with abstinence, since our patients were generally not drinking at enrollment; and misclassification of alcohol intake based on inadequate patient recall of drinking habits and/or oversimplification in our methods of alcohol assessment. Further work is needed to determine how and in whom alcohol contributes to liver disease in HCV.
Women and Liver Disease/Alcohol Intake
Women drank less alcohol than men (median 5.7 g/day vs. 24.1 g/day, P < .0001), and had a lower mean fibrosis score than men (1.39 U vs. 1.62 U, P = .02), although they did not have statistically slower fibrosis progression (0.075 U/year vs. 0.081 U/year, P = .28). The association between alcohol intake and fibrosis was not as clear in women, however. Women did not demonstrate the same stepwise increase in mean fibrosis as men. Only 12% of women drank >50 g/day of alcohol, but this group did not have more fibrosis than the 88% who drank less (OR fibrosis 1.15, P = .72) or compared to nondrinkers (OR fibrosis 0.95, P = .92). Fibrosis scores in women drinking >50 g/day were: Stage 0: 4; Stage 1: 11; Stage 2: 3; Stage 3: 3; Stage 4: 2.
Studies from some years ago found that women experienced toxicity from alcohol at lower doses and had more rapidly progressive alcoholic liver disease than men. More recent studies have generally supported these findings, although some have not. The importance of alcohol intake to liver disease in HCV-positive women, however, has not been reported on extensively. One of the few studies to specifically report alcohol intake by gender in patients with HCV recorded the intake in 57 women only as greater or less than 40 g/day; intake >40 g/day correlated with cirrhosis. In the present study, alcohol intake was not shown to increase fibrosis in 187 women (23% of the cohort). This may be attributable to the fact that alcohol intake in women was very low, with fewer than half drinking more than 4 drinks per week (median: 5.7 g/day). Approximately 35 drinks per week were needed to demonstrate an impact on mean fibrosis overall in this study; even the 23 women who drank this much had a range of fibrosis. Firm conclusions about the risk of alcohol in women with HCV cannot be drawn from our study, but light alcohol consumption was not shown to worsen fibrosis.
ABSTRACT
Heavy alcohol use contributes to liver disease in the setting of chronic hepatitis C virus (HCV) infection. Whether this is true for light or moderate alcohol use has not been demonstrated.
Light alcohol use has survival benefits at a population level and is practiced by most patients with chronic HCV infection. In this study, 800 patients with HCV undergoing liver biopsy at three sites had detailed alcohol histories recorded and the relationship between alcohol and hepatic fibrosis was assessed. On univariate analysis, heavy alcohol use (>50 g/day) was associated with an increase in mean fibrosis (P = .01). Such an association could not be demonstrated for light and moderate alcohol use.
For each category of alcohol intake (none, light, moderate, and heavy), a spectrum of fibrosis was observed. On multivariate analysis, age, serum alanine aminotransferase (ALT), and histological inflammation were the independent predictors of fibrosis (P = <.0001, .0003, <.0001, respectively).
In conclusion, heavy alcohol use exerts a greater effect on fibrosis than light or moderate use. There is a range of fibrosis at each level of alcohol use. Age, serum ALT, and inflammation are independently associated with fibrosis in multivariate analysis, highlighting the fact that variables other than alcohol intake predominate in the production of hepatic fibrosis.
Article Text
INTRODUCTION
Hepatitis C virus (HCV) is the primary cause of cirrhosis leading to liver transplantation in the United States and Europe. Many patients chronically infected with HCV, however, never develop serious liver disease. Epidemiological studies have shown that certain patient-related variables are associated with worse liver disease. In general, these have included: older age at HCV acquisition, heavy alcohol consumption, and male gender. Longer duration of HCV infection, more histological inflammation, and elevated serum liver enzymes have also been found to correlate with liver fibrosis in a number of studies. The proportion of patients progressing to cirrhosis varies widely, based on the group examined, from 22% after 20 years of infection in studies from liver clinics, to 4% in blood donor series. Thus, disease progression clearly depends on a variety of factors.
Heavy alcohol consumption has been found in many studies to contribute to the progression of HCV-related liver disease. Patients with HCV are generally counseled by their physicians to abstain from drinking alcohol, despite the fact that light alcohol intake, which most patients practice, has not been shown to lead to worse liver disease. Additionally, there is increasing evidence that light alcohol consumption bestows significant health benefits. As such, we sought to address how deleterious different amounts of alcohol intake are to patients with chronic HCV.
This issue has not been settled to date, despite many studies that have examined the additive effect of alcohol to HCV-associated liver disease. Each study has had limitations, and none have clearly demonstrated how much alcohol is harmful to the liver. Limitations have included grouping subjects by fixed categories of alcohol intake, with an inability to examine intake within the categories. Also, case-control methodology has often been used, with cases typically having decompensated cirrhosis and controls having little liver disease. Most studies were retrospective, although a few have been prospective.
The contribution of light or even moderate alcohol intake has infrequently been examined. Studies which have examined this issue have had conflicting results, with some finding no relationship between alcohol intake and fibrosis overall, and others suggesting that such a relationship exists. There is no consensus about a level of alcohol that increases the risk of liver disease progression. The specifics of these relationships, however, are important in daily practice for patient counseling.
To overcome these limitations, we conducted a cross-sectional study of alcohol intake up to the time of a liver biopsy. Current and past alcohol use was quantified. Consecutive patients were enrolled at three centers over 5 years. All patients had daily alcohol intake calculated by the same method, allowing the full range to be examined. Histological outcomes were used. Our aims were: 1) to evaluate whether there is a safe level of alcohol intake, and whether alcohol has a dose-effect on fibrosis; 2) to define the spectrum of liver injury in patients who consume the same amount of alcohol; and 3) to compare the contribution of alcohol to fibrosis with that of other common clinical predictors. In doing so, we hoped to clarify the degree to which alcohol contributes to HCV-related liver injury.
RESULTS
Each level of alcohol intake was not associated with significantly increased mean fibrosis compared to the next higher level (i.e., 0 vs. 0.1-20 g/day, 0.1-20 g/day vs. 20.1-50 g/day). Although overall fibrosis was greater in patients with HCV who drink heavily than in those who did not, there was a range of disease in each category of alcohol intake.
AUTHOR DISCUSSION
Intrahepatic inflammation, serum ALT, and patient age were the independent multivariate predictors of hepatic fibrosis in this study. The multivariate odds ratio for inflammation was highly significant, whereas that for heavy alcohol intake just missed being statistically significant. The independent effect of inflammation is supported by La Jolla patients having a similar degree of hepatic fibrosis than patients at the other two sites, despite less alcohol consumption. This could be due to an aggressive immune response in La Jolla patients, supported by their statistically higher ALT. La Jolla patients clearly drank less alcohol than patients from the other two centers, but had similar fibrosis, highlighting the importance of factors other than alcohol in their disease progression.
ALT value and histological inflammation have been found to be associated with fibrosis in several other studies; the inflammatory response to hepatitis C seems to contribute to disease progression. ALT and histological inflammation were highly correlated with one another in this study. The fact that neither ALT nor histological inflammation correlated with alcohol intake in univariate analysis suggests that alcohol does not act through increased inflammation in exerting an effect on fibrosis.
Most studies of risk factors for HCV progression have also found that patient age and duration of HCV infection are primary determinants of histological fibrosis. In univariate models in this study, both age and duration of HCV infection were associated with fibrosis, but in multivariate models only age retained its significance. This may be explained by age being a better surrogate of duration of HCV infection than our estimate of HCV duration, which is based, as in most studies, on time of first injection drug use or blood transfusion. Fibrosis may also be truly an age-dependent process, as some have suggested.
It should be noted that the degree and estimated rate of histological scarring in our cohort differs from some other published studies. Poynard et al., using essentially the same system for scoring fibrosis, found a mean fibrosis score of 1.95 U in their cohort; estimated mean duration of HCV infection was 12.4 years, and the estimated mean rate of fibrosis progression was 0.25 U/year. Our patients had less fibrosis (mean 1.56 U), and a longer duration of HCV (mean 23.7 years), leading to a slower estimated mean rate of fibrosis progression, 0.08 U/year. Our values are similar to those published from other American liver clinic studies. Differences between cohorts likely reflect differences in timing of viral acquisition, host susceptibility, and environmental factors.
|
|
| |
| |
|
|
|
