Editorial: Effect of Atorvastatin, Vitamin E and C on Nonalcoholic Fatty Liver Disease: Is the Combination Required?
Am J Gastroenterol Jan 2011
Bianca M Arendt PhD1 and Johane P Allard MD1
1Department of Medicine, University Health Network, Toronto, Ontario, Canada
Correspondence: Johane P. Allard, MD, Department of Medicine, University Health Network, The Toronto General Hospital, 585 University Ave, Suite 9-NU-973, Toronto, Ontario, Canada M5G 2N2. E-mail: Johane.email@example.com
from original study below-"Univariate predictors of severe to moderate hepatic steatosis at baseline scan include the presence of hypertension (P=0.004), lower high-density lipoprotein (HDL) levels (P=0.0005), higher triglyceride (TG) levels (<0.0001), and higher body mass index (BMI) (<0.0001) (Table 2). We performed multivariate analysis controlling for age, sex, baseline low-density lipoprotein cholesterol, HDL cholesterol and TG levels, systolic and diastolic blood pressures, diabetes, and BMI. Baseline TG levels (odds ratio (OR)=1.003, P<0.001, 95% CI=1.002, 1.006) and BMI (OR=1.10, P<0.001, 95% CI=1.05, 1.14) remained the significant predictors of baseline NAFLD."
FATTY LIVER IN HIV
from Jules: fatty liver in HIV+ is a bigger concern than realized, I think it might emerge as patients age as more of a problem, fatty liver is associated with liver function abnormalities, hepatitis, higher ALTs, so we might expect higher rates of liver disease unrelated to HCV or HBV in aging HIV+ individuals
Nonalcoholic fatty liver disease is associated with the metabolic syndrome. The current standard of care, healthy diet and weight loss, has limited effect. The benefits of pharmacological treatments are unclear due to the difficulty of using liver histology as the main outcome in large randomized controlled trials (RCTs). In this issue, an RCT (1) with atorvastatin and antioxidants (vitamins E+C) vs. placebo shows improvement in liver steatosis based on computed tomography scans. The questions are (1) is this beneficial effect due to the combined treatment or the effect of an individual compound; (2) does this intervention improve nonalcoholic steatohepatitis.
Nonalcoholic fatty liver disease (NAFLD) comprises a disease spectrum from simple hepatic steatosis, which is usually a benign condition, to nonalcoholic steatohepatitis (NASH), which is characterized by inflammation and can lead to fibrosis and cirrhosis (2). NAFLD is associated with central obesity, insulin resistance, dyslipidemia and hypertension (2,3,4), and it is considered the hepatic manifestation of the metabolic syndrome (2).
Current standard of care for NAFLD is weight loss, which-if successful-improves both metabolic parameters and liver histology including inflammation (5). However, in practice, weight loss is often hard to achieve (5,6) and sustainable weight loss requires more complex lifestyle interventions (6,7), or even bariatric surgery (8). Additional treatments were assessed, including antioxidant vitamins (vitamin E, vitamin C), polyunsaturated fatty acids, betaine, or ursodeoxycholic acid (6,7). Most of the data are derived from studies with small numbers of patients, short intervention, and surrogate markers like liver enzymes instead of liver biopsy (7,9). A number of randomized controlled trials (RCTs) showed improvements in liver enzymes and histology with insulin-sensitizing agents, mainly thiazolidinediones (7,10). Recently a multicenter RCT, comparing "Pioglitazone vs. Vitamin E vs. Placebo for Treatment of Non-Diabetic Patients with Nonalcoholic Steatohepatitis" (PIVENS) (11), showed vitamin E (800 IU per day) for 96 weeks improved NASH on liver biopsy compared with placebo, and both pioglitazone and vitamin E were associated with improvements in liver enzymes and reduction in hepatic steatosis (11). Pioglitazone, like other thiazolidinediones, was found associated with weight gain (10,11), which is undesirable in NAFLD.
Beneficial effects of statins on liver enzymes, liver steatosis, and inflammation have been shown in a number of small pilot studies (12,13,14,15,16,17). These benefits were not confirmed in one small RCT examining the effect of 1-year simvastatin treatment on patients with biopsy-proven NASH and hyperlipidemia (18). Only 14 subjects finished the study and 10 had a repeat liver biopsy. Simvastatin reduced serum low-density lipoprotein compared with placebo but there was no statistically significant improvement in serum aminotransferases or liver histology (18). Small sample size may have contributed to the negative results.
The report by Foster et al. (1) in this issue is a large RCT examining the combined effect of atorvastatin (20 mg per day) with antioxidants (vitamin E 1,000 IU per day + vitamin C 1,000 mg per day) on NAFLD assessed by computed tomography (CT). Although the authors state in their abstract that the aim was "to determine if statin therapy may be an effective treatment for NAFLD", truly, we cannot dissect out the effect of atorvastatin from the effect of the antioxidants in this trial which is a substudy of the St Francis Heart Study Randomized Clinical Trial (19). This study included healthy men and women with normal liver enzymes: elevated level of liver enzymes (>1.5 normal) was an exclusion criterion. Excluding patients with elevated level of liver enzymes likely excluded patients with moderate-to-severe NASH. In addition, alcohol intake was not assessed (19), and some participants might actually have had alcoholic rather than nonalcoholic fatty liver.
Foster et al. (1) included in their analysis 455 participants (226 placebo, 229 treatment) who had at least two subsequent CT scans (performed every 2 years) allowing for the calculation of the liver-to-spleen attenuation ratio. With this method moderate-to-severe steatosis (at least 30%) can be identified (1,20). Mean follow-up was 3.6 years. Because the purpose of the initial study was to examine the effect of the treatment on cardiovascular disease, no liver biopsy was performed to assess hepatic inflammation and fibrosis.
The authors identified 80 participants who had NAFLD at baseline. In a multivariate analysis, baseline triglyceride levels and body mass index (BMI) were significant predictors of NAFLD. At the end of the intervention, the proportion of patients with NAFLD in placebo and treatment group was 70.4 and 34.4%, respectively. The odds ratio for NAFLD in the treatment group was 0.29 (95% CI=0.15-0.57, P<0.001). Significant reduction of NAFLD prevalence in the treatment group was still observed in those with normal cholesterol and triglyceride levels at baseline and after controlling for factors like age, sex, baseline BMI and blood pressure, and follow-up lipid values. In the 375 subjects who did not have NAFLD at baseline, 17.6% in the placebo group developed NAFLD compared with 12.1% in the treatment group. This difference was not statistically significant.
These results suggest that the combination of atorvastatin, vitamin E, and vitamin C improves liver steatosis based on CT scan. However, it is not clear whether this is the effect of the combined treatment or atorvastatin alone or vitamin E or C. On the basis of the PIVENS study (11), it is quite possible that vitamin E may have had this effect.
Regarding the combined treatment: (i) the study showed good tolerability with high compliance (85%) over a long treatment period and no major side effects. Transient elevations of transaminase levels occurred only in one subject with NAFLD, suggesting that this combined treatment is tolerable in these patients. (ii) NAFLD patients with normal cholesterol and triglycerides at baseline also responded to treatment, suggesting that dyslipidemia is not necessary for the combined use of atorvastatin and antioxidant vitamins in the treatment of NAFLD, or it may suggest that because antioxidants exert their effects through a different mechanism (reduction in lipid peroxidation), what we are seeing here is more the effect of antioxidants rather than the effect of atorvastatin. (iii) In addition to lipid-lowering effects, statins may have anti-inflammatory properties (21), which could additionally benefit patients with NASH by reducing inflammation and risk of fibrosis. This was not assessed here but was suggested by two small open-label trials in patients with NASH receiving atorvastatin (12) or pravastatin (16). However, in a small RCT, simvastatin did not show any significant effect on liver enzymes, hepatic steatosis, necroinflammatory activity, or stage of fibrosis on liver biopsy (18).
As discussed by the authors (1), the study design has some limitations. The use of a combination therapy does not allow to determine which of the compounds was effective, and if the given doses are ideal. Foster et al. argue that "given the high prevalence of dyslipidemia in patients with NAFLD and the low cost of vitamin E and C, a regimen of the combination therapy would seem to be the best approach." On the basis of the PIVENS study, it is also possible that in those without dyslipidemia, the most active compound from the combined treatment is vitamin E. Considering that (i) vitamin E is the most effective defense mechanism against lipid peroxidation (22); (ii) lipid peroxidation is increased in NAFLD and can promote inflammation and tissue damage (2); (iii) vitamin E requires vitamin C to regenerate (22); (iv) dietary intake of vitamin E and C is frequently lower than the dietary reference range (23,24), it certainly makes sense to recommend to patients with NAFLD to take these supplements. However, we need to keep in mind controversies about the effect of high-dose vitamin E on other outcomes, such as mortality (25).
Finally, as NAFLD is strongly associated with BMI (4), it would have been interesting to report on changes in BMI during the follow-up period and to include these in the multivariate analysis. Furthermore, if the weight of the patients were stable or decreased, this would have shown an advantage of this combined treatment over pioglitazone (11).
Overall, this RCT by Foster et al. (1) is an important contribution to the search for an effective treatment for NAFLD. The authors are using a combination of atorvastatin, vitamin E, and vitamin C, which is affordable and has little side effects. The results are promising and should be the base for trials in NASH involving liver biopsy as end point.
Am J Gastroenterol jan 2011
Atorvastatin and Antioxidants for the Treatment of Nonalcoholic Fatty Liver Disease: The St Francis Heart Study Randomized Clinical Trial - pdf attached
Download the PDF here
Temitope Foster MD, MSCR1, Matthew J Budoff MD2, Sammy Saab MD, MPH, AGAF1, Naser Ahmadi MD2, Craig Gordon BS3 and Alan D Guerci MD3
1Department of Medicine, University of California, Los Angeles, California, USA, 2Los Angeles BioMedical Research Institute at Harbor-UCLA Medical Center, Torrance, California, USA, 3. 3Department of Research, St Francis Hospital, Roslyn, New York, USA
Correspondence: Temitope Foster, MD, MSCR, Department of Medicine, University of California, 10833 Le Conte Avenue Room 44-138, Los Angeles, California 90095, USA. E-mail: firstname.lastname@example.org
OBJECTIVES: Nonalcoholic fatty liver disease (NAFLD) is defined as the spectrum of benign fatty liver to necroinflammation and fibrosis. Its prevalence has been found to be as high as 39%. It is estimated that up to 15% of those affected will go on to have progressive liver disease. Currently, there is no proven therapy for NAFLD. In this study, we aim to determine whether statin therapy may be an effective treatment for NAFLD and identify independent predictors of NAFLD.
METHODS: In all, 1,005 men and women, aged 50-70 years were randomized to receive either a daily combination of atorvastatin 20 mg, vitamin C 1 g, and vitamin E 1,000 IU vs. matching placebo, as part of the St Francis Heart Study randomized clinical trial. Liver to spleen (LS) ratios were calculated on 455 subjects with available computed tomography scans performed at baseline and follow-up to determine NAFLD prevalence. Baseline and final LS ratios were compared within treatment groups, and results were compared between the treatment and placebo groups using univariate and multivariate analyses. Mean duration of follow-up was 3.6 years.
RESULTS: There were 80 patients with NAFLD at baseline. We identified baseline triglyceride levels (odds ratio (OR)=1.003, P<0.001) and body mass index (OR=0.10, P<0.001) as independent correlates of NAFLD. Treatment with atorvastatin combined with vitamins E and C significantly reduced the odds of NAFLD at the end of follow-up, 70 vs. 34% (OR=0.29, P<0.001).
CONCLUSIONS: In conclusion, atorvastatin 20 mg combined with vitamins C and E is effective in reducing the odds of having hepatic steatosis by 71% in healthy individuals with NAFLD at baseline after 4 years of active therapy.
As the rates of obesity continue to rise dramatically in the United States (1), so do the associated comorbidities. Nonalcoholic fatty liver disease (NAFLD) is defined as the spectrum of benign fatty liver infiltration, or simple hepatic steatosis, to necroinflammation and fibrosis. The prevalence of NAFLD in the United States has not been well described but has been found to range from 15 to 39% (2). The majority of these patients have a benign fatty infiltrate, which is considered to be nonprogressive. However, it is estimated that up to 10-25% of those with NAFLD (or 2-7% of the entire population) will have nonalcoholic steatohepatitis (NASH), a combination of necroinflammation and fibrosis, which can ultimately lead to cirrhosis (3,4).
NAFLD is widely accepted to be the hepatic manifestation of metabolic syndrome, a known cause of increased mortality (5). Metabolic syndrome is characterized by a group of metabolic risk factors, including central obesity, dyslipidemia, insulin resistance, a prothrombotic state (e.g., high fibrinogen), high blood pressure, and a pro inflammatory state (e.g., elevated C-reactive protein) (6). The exact prevalence of NAFLD in patients with metabolic syndrome is unknown, though a high proportion of those with metabolic syndrome have elevated aminotransferases (7). However, 36% of patients with NAFLD have 3 or more components of metabolic syndrome and this number increases to 67% in those who are obese (5). NAFLD is now also known to independently increase overall mortality, with malignancy and cardiovascular disease being the most common cause of death (8,9).
Presently, we have no proven effective therapy for NAFLD. A large majority of the studies looking at potential therapies have been open-label pilot studies with small sample sizes (10). Treatment with antioxidants, insulin sensitizers, and weight loss, either through diet or weight-loss surgery, has shown some promise, although most of these data are inconclusive (11,12,13,14,15,16,17). Approximately 70% of patients with NASH also have concurrent dyslipidemia (2) making treatment with a lipid-lowering medication appear to be a reasonable approach. A few small pilot studies have noted atorvastatin, a 3-hydroxy-3-methylglutaryl-coenzyme A reductase inhibitor, to be effective. A study evaluating 31 patients treated with 24 months of atorvastatin showed improvement in their NAFLD activity score and transaminase levels (2). Another distinct study had similar findings in 10 patients with NAFLD and dyslipidemia treated with atorvastatin for a mean of 38 weeks (18).
Given the increased mortality rates in patients with NAFLD, the rising rates of obesity, diabetes, and metabolic syndrome in this population, finding an effective therapy is of utmost importance. In this study, our aim is to determine whether statin therapy may be an effective treatment for NAFLD based on radiographic evidence, as well as identify independent predictors of NAFLD in an otherwise healthy population.
A total of 455 subjects were analyzed: 226 in the placebo group and 229 in the treatment group. Roughly 93% of the population was non-Hispanic white, with a mean age of 59 (s.d.± 6.0) years. Medication compliance from the original cohort was assessed every 3 months and was defined as consumption of at least 85% of the study medication. The subjects averaged 85% compliance for the active therapy and its matching placebo. The average length of follow-up was 3.6± 1.1 years. Baseline characteristics were similar between groups and are listed in Table 1.
Univariate predictors of severe to moderate hepatic steatosis at baseline scan include the presence of hypertension (P=0.004), lower high-density lipoprotein (HDL) levels (P=0.0005), higher triglyceride (TG) levels (<0.0001), and higher body mass index (BMI) (<0.0001) (Table 2). We performed multivariate analysis controlling for age, sex, baseline low-density lipoprotein cholesterol, HDL cholesterol and TG levels, systolic and diastolic blood pressures, diabetes, and BMI. Baseline TG levels (odds ratio (OR)=1.003, P<0.001, 95% CI=1.002, 1.006) and BMI (OR=1.10, P<0.001, 95% CI=1.05, 1.14) remained the significant predictors of baseline NAFLD.
NAFLD response to active therapy
There were a total of 80 patients with LS ratios <1 on CT: 36 in the placebo group and 44 in the treatment group. By year 2, this population was reduced to 62: 27 in the placebo group and 35 in the treatment group. At year 4, the total population was 59: 27 in the placebo group and 32 in the treatment group. Baseline characteristics were similar between groups, with the exception of the presence of more subjects with a family history of cardiovascular disease in the placebo group in comparison with the treatment group (P=0.05) (Table 3).
Among the 80 patients with NAFLD, total cholesterol and low-density lipoprotein cholesterol levels were significantly reduced by treatment: 210.6± 31.6 (control) vs. 169.9± 38.6 (treatment, P<0.001) and 123.3± 39.9 (control) vs. 90.3± 37.9 (treatment, P<0.001), respectively. The follow-up mean HDL values; 46.8± 16.9 (placebo) vs. 43.5± 8.3 (treatment, P=0.25), and mean TG levels; 195.4± 131.6 (placebo) vs. 169.4± 123.4 (treatment, P=0.37), did not differ significantly between treatment and control groups.
There was a higher percentage of patients with NAFLD at the time of the second CT scan (2.1± 1.0 years) in the placebo group (77.8%) when compared with the treatment group (37.1%, P=0.001). At the end of follow-up (3.6± 1.1 years) this difference remained significant; 70.4% (placebo) vs. 34.4% (treatment; OR=0.29, 95% CI=0.15-0.57, P<0.001) (Figure 1). The average LS ratio increased from 0.71± 0.40 to 1.106± 0.52 in subjects assigned to treatment with atorvastatin (Figure 2). This difference was still significant after controlling for baseline HDL, low-density lipoprotein, and TGs (OR=0.28, 95% CI=0.14-0.57, P<001). In subgroup analyses of those with a total cholesterol <200 (n=21, OR=0.12, 95% CI=0.02-0.90, P=0.04) and TG level <150 (n=34 OR=0.17, 95% CI=0.06-0.51, P=0.002), there was still an improvement in NAFLD seen in those on active therapy. When we controlled for age, sex, baseline BMI, race, baseline systolic and diastolic blood pressure, baseline diabetes, and follow-up lipid values, there was still a significant reduction in the prevalence of fatty liver in the treatment group (OR=0.36, 95% CI=0.16-0.83, P=0.017). Follow-up TG levels were a significant predictor of NALFD at the end of treatment (OR=1.004, 95% CI=1.004-1.0074, P=0.02), as well as systolic blood pressure (OR=0.95, 95% CI=0.93-0.99, P=0.007) in this model.
Atorvastatin, vitamins E and C, and NAFLD development
In those without NAFLD at baseline (n=375), we found that 17.6% of the placebo group developed NAFLD compared to 12.1% of the treatment group. This difference was not statistically different (P=0.23).
There were three subjects in our study with transaminase elevations greater than twice the upper laboratory limit of normal by year 2 of follow-up. These elevations were all resolved by year 4 of follow-up (Table 4).
In this study, we evaluated the effects of 4 years of 20 mg daily atorvastatin treatment combined with daily vitamins E and C, in apparently healthy individuals with radiographically diagnosed NAFLD, i.e., hepatic steatosis >30%. There were a total of 455 subjects included in this study, 80 of whom had NAFLD at baseline. There was a 71% reduction in the risk of having moderate-to-severe hepatic steatosis at the end of the study in those on treatment. In order to evaluate whether the reduced prevalence of NAFLD at the end of the study was a direct effect of treating and improving dyslipidemia, we performed a subgroup analysis on NAFLD patients with baseline total cholesterol values <200 and/or TG levels <150. We found that the risk reduction remained significant, although this population was smaller and so further studies are necessary for definitive conclusions to be drawn. Follow-up TG levels were a significant predictor of fatty liver at the end of the study among those with NAFLD at baseline, with an OR of 1.004. This estimated odds ratio would imply that a 50 unit increase in triglyceride levels increases the odds of continuing to have NAFLD by 20%. Follow-up SBP was also a significant predictor of NAFLD at the end of the study, with an OR of 1.007 i.e. at 10 point increase in SBP increases the odds of NAFLD at follow-up by 7%.
We found that in patients without NAFLD at baseline, more people in the placebo group (n=22) developed fatty liver than those receiving active therapy (n=14). This difference was not statistically different (P=0.23), possibly because of the overall small numbers. Further studies must be carried out in order to draw any significant conclusions on the impact this combination of drugs has on preventing the development of NAFLD.
Similar to other studies, we found that hypertension, low HDL, and increased TG levels (all components of the metabolic syndrome) were significant correlates of baseline fatty liver in this population (6). Interestingly, we did not find a significant association with diabetes and baseline NAFLD. This is likely due to the overall low number of diabetics in this population (n=7).
This study showed a significant reduction in moderate-to-severe hepatic steatosis in a randomized population treated with a combination of atorvastatin, vitamins E and C. Our findings also confirm the evidence obtained from pilot data showing the efficacy of statin therapy on NAFLD (Table 5). Approximately 70% of patients with NASH also have been found to have concurrent dyslipidemia (2). Another distinct study found 60% of patients with mixed hyperlipidemia had evidence of hepatic steatosis (25), making atorvastatin an ideal treatment. We have shown atorvastatin to be beneficial in those with normal cholesterol (<200; OR=0.12, 95% CI=0.02-0.90) and TG (<150; OR=0.17, 95% CI=0.06-0.51) levels, making it a potential therapy in those NAFLD patients without dyslipidemia. The evidence on the efficacy of vitamin E has been more equivocal; although some studies have shown vitamins E and C to be a promising treatment of NAFLD (15,26), others have shown them to be ineffective (16,27). However, recent unpublished data from the PIVENS (pioglitazone versus vitamin E for the treatment of non-diabetic patients with non-alcoholic steatohepatitis) trial (28) did show a significant improvement in NASH in those taking vitamin E when compared with pioglitazone and placebo. This has been the largest randomized trial to date to show the efficacy of vitamin E, making it a viable treatment option.
The potential for hepatic toxicity from statins, ranging from transaminase elevations to the rare occurrence of acute liver disease, remain a concern for many practitioners. Systematic reviews have yielded variable results; although some have found no difference in the incidence of transaminase elevations in patients receiving moderately dosed statins (29), other studies have found a positive relationship between statin dose and transaminase levels (30). At this time, however, the implications of transaminase elevations caused by statin therapy are not well known and have not been well correlated to the degree of histological injury (31). In our study, there were only three patients who developed transaminase elevations greater than two times the upper laboratory normal value in year 2. Of the two patients with four year follow-up transaminases available, patient 3's values had completely normalized and patient 1's values had significantly decreased.
This study does have a few limitations. First, our active treatment is a cocktail of atorvastatin, vitamins C and E. It is difficult to determine the extent each of the treatment components contributed to the treatment results or if the results were due to synergy. Given the high prevalence of dyslipidemia in patients with NAFLD and the low cost of vitamins E and C, a regimen of the combination therapy would seem to be the best approach. Second, our measurement of hepatic steatosis, although objective, is not the gold standard. Currently, we do not have noninvasive means to determine which patients with fatty liver infiltrate have NASH. We are also limited in our ability to predict which patients with fatty liver will go on to develop NASH. This is challenging for a number of reasons: we cannot perform invasive liver biopsies on every patient with fatty liver infiltrate, yet an estimated 7-17% of those patients with NASH are thought to go on to develop cirrhosis (4,6). CT scans are not able to accurately characterize the amount of hepatic fat when the total content is <30%. Hepatic steatosis >30%, however, is diagnosed fairly accurately with CT using the LS ratio, with the area under the receiver operating characteristic curve of 0.991 (24). Therefore, we were able to identify those with moderate-to-severe hepatic steatosis and hope that this is an adequate surrogate for those who also have NASH. Third, although patients with transaminase levels >1.5 times the upper limit of normal were excluded, specific cases of chronic liver disease were not identified for exclusion. With the relatively low prevalence of these diseases in the general population, the impact on this trial is likely low and any possible confounding should have affected the treatment and placebo groups equally. Finally, we excluded all those without both a liver and spleen on their CT scan. It is unclear, though it would seem unlikely that these scans were incomplete due to patient level characteristics, these data were likely missing at random and was treated as such.
In conclusion, atorvastatin 20 mg in combination with vitamins C and E lowered the risk of having moderate-to-severe hepatic steatosis by 70% in a healthy population of 80 patients with NAFLD at baseline after 4 years of therapy. We also have convincing evidence that it is equally efficacious in patients without dyslipidemia, although further studies with a larger population need to be conducted.