The primary aim of the study is to determine the impact on hepatic steatosis of replacing premeal rapid-acting insulin for exenatide (Byetta) while maintaining bedtime long-acting detemir (Levemir) insulin in well-controlled patients with type 2 diabetes mellitus (T2DM) and nonalcoholic fatty liver disease (NAFLD).

Secondary aims are to learn: 1) the efficacy and safety of such approach and whether it is an acceptable treatment strategy compared to intensified insulin therapy alone; 2) mechanisms of action (effects on insulin secretion and insulin action); 3) its impact on weight (can it prevent insulin-associated weight gain or cause weight loss) and rates of hypoglycemia; 4) if it may improve specific plasma biomarkers of disease activity in NAFLD and inflammatory markers common to both conditions - T2DM and NAFLD (hsCRP, ICAM, VCAM).

Type 2 diabetes mellitus (T2DM) is a major public health problem in the United States with ~2/3 of Americans that are overweight or frankly obese. Less well recognized is that obesity and T2DM are fueling another "silent epidemic": non-alcoholic fatty liver disease (NAFLD). In NAFLD, hepatic fat accumulation ranges from simple steatosis to severe steatohepatitis (NASH), with necroinflammation and fibrosis that may progress to cirrhosis. Indeed, up to 40% of patients with NAFLD develop NASH and about ~15-20% of patients with NASH develop cirrhosis after 10 years. NASH is believed to be the most common cause of cryptogenic cirrhosis and ranking as the third leading indication for liver transplantation after alcohol and hepatitis C. The distinction between simple steatosis and NASH can only be done by means of a liver biopsy, but strategies that reduce fatty liver disease correlate with a reduction in steatohepatitis and fibrosis.

It is estimated that ~1/3 of the general population of the United States have fatty liver disease as assessed by magnetic resonance imaging and spectroscopy (MRS), the gold-standard technique. It is also affecting the pediatric population, and while the natural history remains to be defined, most biopsies show some degree of fibrosis and cirrhosis has been reported in children as young as 10 years of age. Moreover, in the investigators experience using MRS for the past several years, ~80% of patients with T2DM have NAFLD. If the above is extrapolated to the overall population of the United States, there are ~15 million subjects with T2DM and NAFLD and more than 50 million overweight/obese individuals with NAFLD.

As awareness of NAFLD/NASH increases, health care providers are now confronted with the unique challenge of achieving good metabolic control while treating fatty liver disease in the same subject. Inadequate glycemic control appears to worsen and accelerate progression of fatty liver disease. Paradoxically, most patients with T2DM do not reach established treatment goals. Moreover, it appears that the way metabolic control achieved is important in NAFLD: while metformin may be beneficial in NASH and pioglitazone is highly effective, control of hyperglycemia by insulin therapy alone has modest benefit on hepatic fat accumulation. In addition, long-term insulin therapy in T2DM is frequently associated with inadequate compliance, weight gain, frequent hypoglycemia and overall poor patient satisfaction. This is of great concern as ~35-40% of patients with T2DM are treated with insulin. Therefore, there is an unmet need to find better ways to treat patients that fail oral agent therapy, targeting both hyperglycemia and excessive liver fatty deposition.

The investigators propose a strategy of adding exenatide to insulin therapy as a means to achieve better glycemic control while ameliorating hepatic steatosis (and improving hepatic insulin sensitivity), reduce body weight/total body fat (and visceral fat), decrease the risk of hypoglycemia by allowing a reduction in insulin doses (withdrawal of premeal insulin doses, reduction of basal long-acting insulin), and enhance insulin secretion and insulin action while improving the quality of life of patients with T2DM and NAFLD. As both T2DM and NAFLD will become more prevalent in the near future, this pilot study will likely set the stage for a large controlled multicenter trial with an approach that may be more effective than intensified insulin therapy alone.

The central hypothesis is that insulin resistance and fat-overload play a key role in NAFLD/NASH, a condition believed to arise at the molecular level from the inability of the mitochondria to adapt to fat oversupply. Excessive fatty acid flux from exogenous (dietary) and endogenous (insulin resistant adipose tissue/increased lipolysis) sources drive hepatic lipogenesis. Exenatide would potentially be a good treatment fit against hepatic steatosis by decreasing excess dietary intake and substrate supply, reducing fat mass and plasma FFA and by lowering exogenous insulin needs and chronic hyperinsulinemia. Mitigating chronic hyperinsulinemia and hyperglycemia is important as they drive hepatic fat synthesis through sterol regulatory element-binding protein 1c (SREBP-1c) and carbohydrate regulatory element-binding protein (ChREBP) activity, respectively. Weight loss may also subdue systemic inflammation generated by dysfunctional adipose tissue, and at the local (hepatic) level, ameliorate fat-induced mitochondrial dysfunction (activation of Kupffer cells, local production of cytokines).

While no systematic studies have been published yet, exenatide has been shown to reverse hepatic steatosis in an obese animal model of NAFLD and in a case report in a 59 year old poorly controlled T2DM patient in whom liver fat by MRS was reduced from 15.8% to 4.3%. Dietary intervention remains as the current standard of care for NAFLD, but studies in general have been small, uncontrolled and led to variable histological results. It is widely accepted that weight loss is difficult to achieve and even harder to maintain in the long-term. Moreover, weight loss is particularly difficult in patients with T2DM on insulin therapy. Pharmacological therapy at large has been ineffective in NAFLD/NASH, including trials using antioxidants and cytoprotective agents such as pentoxyfilline, vitamin E and ursodeoxycholic acid. A modest benefit has been reported in small studies with metformin, but only our study using pioglitazone has shown in a randomized, placebo-controlled trial, to be truly effective and safe in the treatment of patients with IGT or T2DM and NASH.

Considering the impact that NAFLD has on patients with T2DM as a serious co-morbidity ranging from its metabolic impact on glycemia and dyslipidemia, to potentially causing end-stage liver disease and cardiovascular disease, it is rather surprising that no prior studies have focused on novel pharmacological approaches such as the one proposed for the treatment of patients with NAFLD and T2DM.

Inclusion Criteria:

To enter the study subjects must meet the following criteria:

1. Have been on intensified insulin therapy with insulin detemir (Levemir) and premeal insulin aspart (Novolog®) for the previous 6 months.
2. Be able to communicate meaningfully with the Investigator and be legally competent to provide written informed consent.
3. Female patients must be non-lactating and must either be at least two years post-menopausal, or be using adequate contraceptive precautions (i.e. oral contraceptives, approved hormonal implant, intrauterine device, diaphragm with spermicide, condom with spermicide), or be surgically sterilized (i.e. bilateral tubal ligation, bilateral oophorectomy). Female patients who have undergone a hysterectomy are eligible for participation in the study. Female patients (except for those patients who have undergone a hysterectomy or a bilateral oophorectomy) are eligible only if they have a negative pregnancy test throughout the study period.
4. Age range of 18 to 70 years (inclusive).
5. Patients must have been on a stable dose of allowed chronic medications for 6 months prior to entering the double-blind treatment period.
6. All participants must have the following laboratory values:

Hemoglobin ≥12 g/dl in males or ≥11 g/dl in females Serum creatinine ≤1.5 mg/dl AST (SGOT) and ALT (SGPT) ≤2.5 times upper limit of normal Alkaline phosphatase ≤2.5 times upper limit of normal

Exclusion Criteria:

Patients will be excluded if any of the following criteria are present:

1. Individuals with type 1 diabetes or type 2 diabetes and a FPG ≥ 300 mg/dl; poor compliance with insulin therapy.
2. Subjects on sulfonylureas, metformin and/or TZDs unless the dose has been stable for at least 6 months prior to study entry.
3. Patients on any of the following medications: thiazide or furosemide diuretics, beta-blockers, or other chronic medications with known adverse effects on glucose tolerance levels unless the patient has been on stable doses of such agents for the past two months before entry into the study. Patients may be taking stable doses of estrogens or other hormonal replacement therapy if the patient has been on these agents for the prior two months. Patients taking systemic glucocorticoids will be excluded.
4. Past (within 1 year) or current history of alcohol abuse.
5. Patients will be excluded if there is a history of clinically significant heart disease (New York Heart Classification greater than grade II), peripheral vascular disease (history of claudication), or pulmonary disease (dyspnea on exertion of one flight or less; abnormal breath sounds on auscultation) or chronic renal failure (serum creatinine greater than 1.5 mg/dl).