Sirs, We read with interest the report of the effect of atorvastatin treatment on non-alcoholic fatty liver disease (NAFLD) indices in dyslipidaemic patients (n =22).1 Atorvastatin (10–80 mg daily) administered for 6–12 months normalized transaminase activity in 56% of the participants without any change in liver fat content (as assessed by ultrasonography).1 Indeed, two small pilot studies had shown improvement of liver enzyme activities with atorvastatin.2, 3 Furthermore, pravastatin (20 mg daily for 6 months) normalized liver enzymes and improved hepatic inflammation in five patients with non-alcoholic steatohepatitis.4 Statins were also safe when administered to subjects with elevated baseline transaminases.5, 6
We evaluated the effect of a multifaceted approach on NAFLD indices in subjects (n = 186) with the metabolic syndrome and both biochemical and ultrasonographic evidence of NAFLD.7 Patients received lifestyle advice and treatment for hypertension (mainly inhibitors of the renin-angiotensin system), impaired fasting glucose (metformin), obesity (orlistat) and dyslipidaemia [randomly allocated to atorvastatin 20 mg/day (n = 63) or micronized fenofibrate 200 mg/day (n = 62) or both drugs (n = 61)] for 12 months. At the end of treatment, 67% of patients on atorvastatin (as well as 42% on fenofibrate and 70% on combination treatment) no longer had both biochemical plus ultrasonographic evidence of NAFLD (P < 0.05 vs. baseline for all comparisons).7 In addition, 11% of subjects in the atorvastatin group had an echogenic liver at the end of the study but normal transaminases. Therefore, transaminases normalized in 78% of subjects allocated to atorvastatin compared with 56% in the study by Gomez-Dominguez et al.1 How can we account for this discrepancy together with the lack of ultrasonographic regression in the study by Gomez-Dominguez et al?1 One possibility is that the benefit of atorvastatin in our study was seen ‘on the top’ of multifactorial treatment aimed at tackling all components of the metabolic syndrome.8 For example, body weight fell by 12% in the atorvastatin arm in our study (P < 0.05 vs. baseline)7 but did not significantly change in the study by Gomez-Dominguez et al.1 Even modest weight loss alone is associated with normalization of transaminases.9
In conclusion, statin treatment is a promising option for the management of patients with NAFLD who also have dyslipidaemia.10 However, randomized placebo-controlled studies are needed to document a biopsy-proven improvement before we can consider statins as a specific treatment for NAFLD. Furthermore, statins may work better within a multifactorial management of all metabolic risk factors associated with NAFLD.