Is There A Role For Insulin Sensitizers In The Treatment Of NASH? Claims And Controversy


Written By

Christine Choy, PharmD, BCPS

Reviewed By

Dawn Havrda, PharmD, BCPS
Stuart Haines, PharmD, BCPS, BC-ADM


Sanyal AJ, Chalsani N, Kowdley KV, et al. Pioglitazone, Vitamin E, or Placebo for Nonalcoholic Steatohepatitis. N Engl J Med. 2010;363:1675-85.

Non-alcoholic fatty liver disease (NAFLD) is a condition characterized by fat deposition in the hepatocytes of patients with minimal or no alcohol intake.  NAFLD has emerged as the leading cause of chronic liver disease in the Western world, with an estimated prevalence of 20-30% in the general population.1  Patients with NAFLD may present along spectrum of histological findings, ranging from simple hepatic steatosis (fat accumulation without inflammation or fibrosis) to nonalcoholic steatohepatitis (NASH).   Among patients with NASH, 15-20 percent may progress to cirrhosis secondary to chronic hepatocellular inflammation and necrosis.2-4  Furthermore, there is alarming evidence that suggests NAFLD and NASH may be associated with fatal and nonfatal cardiovascular disease (CVD) events5. 

The precise etiology of NASH is unknown, but it is strongly associated with obesity, insulin resistance, and dyslipidemia.6  Data to support pharmacotherapeutic treatments for NASH are sparse, but there is interest in studying the impact of insulin sensitizing therapy on the disease based on its association with components of the metabolic syndrome. In particular, the recently published PIVENS study in the New England Journal of Medicine has caused some buzz about whether patients with NASH should received pioglitazone or high-dose vitamin E. 7  The short answer is no.  But let’s look at the data.

The Pioglitazone, Vitamin E, or Placebo for Nonalcoholic Steatohepatitis (PIVENS) trial was a phase 3, multicenter, randomized controlled trial that recruited patients with a probable or definite diagnosis of NASH based on a histologic disease activity scoring system.  Patients were excluded if they consumed significant amounts of alcohol (i.e., two drinks daily for women or three drinks daily for men) or had a concurrent diagnosis of diabetes, cirrhosis, hepatitis, or symptomatic heart failure.   Two hundred forty seven adults were randomized to pioglitazone 30 mg daily, vitamin E 800 IU daily, or placebo.  The primary outcome was improvement in histological findings – including a composite of scores for hepatocellular ballooning, fibrosis, and NAFLD activity score.  Secondary outcomes included the individual components of the composite outcome.  Two hundred forty patients were required to achieve 90% power, and a P< 0.025 was required to demonstrate a significant difference between groups. 

After two years of follow-up, significantly more patients in the vitamin E group had an improvement in the primary outcome compared to placebo (43% vs 19%, P=0.001), with a number needed to treat (NNT) of 4.  Improvements were noted for all individual components of the histological score, including steatosis, lobular inflammation, hepatocellular ballooning, and NAFLD activity score.     In contrast, there appeared to be no difference between pioglitazone and placebo in the primary outcome rate (34% vs 19%, P-0.04), even though pioglitazone was significantly better than placebo in the majority of individual components of the histological score.  Does this mean that pioglitazone therapy should be discarded as a therapeutic option?  Not necessarily.  It is likely that pioglitazone failed to demonstrate a significant benefit in the primary outcome for one simple reason.    Randomization was not achieved with respect to the percentage of patients with and without heptatocellular ballooning at baseline.   In particular, more patients in the pioglitazone group did NOT have hepatocellular ballooning at baseline (17% placebo, 18% vitamin E, and 28% pioglitazone).   Therefore, it may have been harder for the pioglitazone group to demonstrate a significant benefit in this endpoint. 

Do the potential benefits of vitamin E and pioglitazone outweigh the risks?  With respect to adverse events, there was significantly more weight gain in the pioglitazone group compared with placebo (4.7 kg vs 0.4 kg).  No differences were noted between either treatment group and placebo with respect to CV events, heart failure, bone fracture, severe hepatotoxicity, or progression to cirrhosis.   Unfortunately, the small study population and short duration of follow-up limits interpretation of the adverse event data.  Furthermore, there is known data about the potential mortality risk associated with high dose vitamin E therapy (i.e., greater than 400 IU daily).8

Given the prevalence of the metabolic syndrome and the strong association between the metabolic syndrome and NAFLD, I am surprised that more patients do not present with a diagnosis of NASH in their past medical history.  I suspect the disease is underdiagnosed due to reluctance to confirm diagnosis with an invasive liver biopsy when treatment options are limited.   At present, it is unclear if the benefits of thiazolidinedione therapy for the treatment of NASH outweigh the risks.    The PIVENS trial provides the most compelling evidence to date that pioglitazone has a positive impact on clinical markers of the disease.  But we need more data about the impact of insulin sensitizers on clinical outcomes of NASH, including cirrhosis, end-stage liver disease.  Furthermore, the study was underpowered and too short in duration to detect a clinically meaningful difference in CV disease events and heart failure.  Recent meta-analyses may mitigate fears about the association between pioglitazone and cardiovascular risk.9-10  However, the literature continues to warn about the risk of edema and heart failure associated with long-term thiazolidinedione therapy.  For example, both PROACTIV and DREAM, demonstrated a significant increase in new onset heart failure associated with thiazolidinedione use – in patients with prediabetes and diabetes.11-12  Finally, the risk of bone fracture must be evaluated because patients with liver disease are inherently at higher risk.  Future trials of insulin sensitizers in patients with NASH should critically evaluate the risk of edema, heart failure, and fracture in order to assess these safety concerns. 

What about other insulin sensitizing agents?  Preliminary studies evaluating metformin use in patients with NASH show promising data, including reductions in ALT concentrations and improvement in echogenic liver response.13  But to date, trials of metformin only evaluated nonspecific markers of NASH and definitive benefit has not been shown.  Further study is also required to elucidate risks, especially of metformin-associated lactic acidosis in patients with liver disease.  In the interim, risk factor modification remains the mainstay of treatment.  Weight loss should be advocated, as there is compelling evidence that a 7% reduction in weight results in histological improvements in patients with NASH.14 Bariatic surgery also appears to benefit NASH-outcomes.15  However, at present, there is insufficient evidence to recommend pioglitazone for most patients with NASH.  Additional controlled trials are critical to elucidate the benefits and risks of insulin sensitizers in this population.

What do you think?  Is there compelling evidence to support use of insulin sensitizers, like pioglitazone or metformin, for the treatment of NASH?