Difficult recruitment and end of funding.

Background: Non-alcoholic fatty liver disease (NAFLD) is a spectrum of disorders characterized by lipid accumulation in hepatocytes. Evidence shows that thyroid hormone might be beneficial for this condition.

Objective: To determine whether low dose levothyroxine (LT4) therapy may be a potential treatment for diabetic patients with NAFLD in a single arm study.

Primary: To ascertain whether administration of LT4 for 16 weeks by titrating the serum thyroid stimulating hormone (TSH) to 0.34 mIU/L - 1.7 mIU /L reduces liver fat content by at least 3% among patients with type II diabetes as measured by functional MRI.

Secondary: To ascertain whether administration of LT4 for 16 weeks by titrating the serum TSH to 0.34 mIU/L - 1.7 mIU /L can improve glycemic control as measured by reduction in glycosylated hemoglobin (HbA1c), improve serum lipid profile in Type II diabetic patients with NAFLD as measured by total serum cholesterol, high-density lipoprotein (HDL), low-density lipoprotein (LDL) and total triglycerides (TG) and reduce the proportion of liver fat over body fat, which is reflected by fat in abdominal subcutaneous and visceral tissues, as measured by functional MRI on abdomen.

Subjects and Centres: A total of 50 eligible adult diabetic men with NAFLD will be recruited from 6 centres in Singapore - Changi General Hospital (CGH), Singapore General Hospital (SGH), Tan Tock Seng Hospital (TTSH), National University Health System (NUHS), Khoo Teck Puat Hospital (KTPH), Jurong Health (JH)

Eligible patients: Males between 21 to 60 years of age diagnosed with stable Type II diabetes mellitus (DM) with a baseline alanine aminotransferase (ALT) < 3 times upper limit of normal as per the institution's specified reference range, with a liver ultrasound (US) showing presence of fatty liver and baseline Thyroid stimulating hormone (TSH) levels between 1 - 10 mIU/L.

Treatment: Low dose levothyroxine (LT4) for 16 weeks, not including the 12 weeks of pre-study titration of LT4 in order to attain target TSH level of 0.34-1.70 mIU/L.

Statistical Analysis: The absolute change in liver fat content from baseline (primary endpoint) will be analyzed using one-sample two-sided t-test at a 5% significance level. The same test will be applied to secondary endpoints. Mean, standard deviation and 95% confidence interval will be calculated for primary endpoint and secondary endpoints.

Background: Non-alcoholic fatty liver disease (NAFLD) is a spectrum of disorders characterized by lipid accumulation in hepatocytes. These conditions range from fat accumulation in the liver (hepatosteatosis) to non-alcoholic steatohepatitis (NASH), which in turn, increases the risk for cirrhosis and hepatocellular cancer. NAFLD occurs in 25-35 % of the general United States population, and its prevalence is estimated to be 60 %-80 % for patients with type II diabetes mellitus (DM). Its co-occurrence with type II DM significantly increases the risks for renal and cardiovascular disease. The prevalence of NAFLD in Singapore is not known. However, given that the estimated percentage of adults with type II DM or obesity are approaching 10 % and rapidly rising, a significant proportion of the Singaporean population is likely at risk for NAFLD. Although there are drug therapies for diabetes; currently, there is no effective treatment for NAFLD.

Although NAFLD may cause inflammation, lipotoxicity, and scarring within the liver, it also can have profound metabolic effects as it induces hepatic insulin resistance. Defects in β-oxidation of fatty acids are thought to play an important role in the pathogenesis of NAFLD, which then leads to further insulin resistance and worsened glycemic control in diabetic patients. Since thyroid hormones (THs: T3, T4) promote the oxidation of fatty acids within the liver, it is possible that ineffective TH action in the liver may contribute to NAFLD.

It has been shown recently that the incidence of hypothyroidism is doubled in patients with NAFLD as approximately 15% patients are affected. Additionally, several TH and TH analogues are able to ameliorate NAFLD in rodents fed with high fat chow diets. At the genomic level, many of the genes that have altered expression in NAFLD are regulated by TH further supporting the notion that defects in TH signalling may promote hepatosteatosis and hepatic damage. Two recent studies also have shown thyroid hypo-function in both young and elderly adults with NAFLD.

In this study, we seek to examine whether short-term low dose levothyroxine can decrease hepatic fat deposition and improve glucose control in diabetic patients with NAFLD.

Accordingly, we plan to treat 50 male patients with stable Type II DM and NAFLD with low dose LT4, and assess fat content in the liver by MRI and glucose control by measuring serum fasting glucose, HbA1c, and degree of insulin resistance as determined by the homeostatic model assessment (HOMA-IR). The potential benefit for the patient will be reduction of hepatic fat content and/or improvement in glucose control. The potential major side effects are atrial arrhythmias and osteoporosis. The risk for the latter side effects will be minimized by treating subjects that do not have any known heart disease or osteoporosis with low dose levothyroxine. It is expected that thyroid hormone levels will be in the high normal range. Patient will undergo hepatic MRI at beginning of study and at 16 weeks and phlebotomy at beginning of study, at 8 weeks, and at 16 weeks. The study will be conducted in compliance with the protocol, Singapore Guideline for Good Clinical Practice (SGGCP), site's respective Institutional Review Board (IRB), and Health Science Authority (HSA) requirements.

Preliminary Studies:To investigate the potential etiological role of TH in hepatosteatosis, we initiated studies in which human liver HepG2 cells over-expressing TH receptors (TRs) were treated with palmitate in the absence or presence of T3. In this cell culture model of NAFLD, the uptake of palmitate and the intracellular accumulation of fatty acids caused apoptosis in the absence of hormone. Surprisingly, T3 treatment blocked the palmitate-induced apoptosis. Additionally, our studies suggest that part of this protective effect by T3 is due to induction of "lipophagy", a type of autophagy in which cells consume and digest fatty acids before delivery to mitochondria. Concomitantly, we have observed induction of cpt1a mRNA, a protein which transports fatty acids to mitochondria. Our findings suggest that T3 may facilitate both the consumption and β oxidation of fatty acids within the hepatocyte. If this is the case, and when taken together with previous findings, our data would suggest the liver may be particularly susceptible to lipid accumulation and damage in hypothyroidism. Additionally, NAFLD itself may engender a "relative hypothyroid" state within the liver - resulting in impaired responsiveness of key metabolic target genes to TH. Thus, it is possible that TH, or TH analogues may be useful in the treatment and prevention of NAFLD.

Our cell culture and in vivo data, in combination with previous studies in rodents showing that T3 and its analogues may be able to decrease hepatosteatosis and evidence that TH action may be decreased in human liver in obese patients undergoing gastric bypass and in patients with NAFLD, provide the experimental rationale to examine whether short-term levothyroxine therapy can decrease hepatosteatosis in Type II diabetic patients and improve glycemic control in Type II diabetic patients with NAFLD.

Objective: To determine whether low dose levothyroxine (LT4) therapy may be a potential treatment for diabetic patients with NAFLD.

Primary Objective:

- To ascertain whether administration of LT4 for 16 weeks by titrating the serum TSH to 0.34 mIU/L - 1.7 mIU /L reduces liver fat content by at least 3% among patients with type II DM as measured by functional MRI.

Secondary Objective:

• To ascertain whether administration of LT4 for 16 weeks by titrating the serum TSH to 0.34 mIU/L - 1.7 mIU /L can improve glycemic control in Type II diabetic patients with NAFLD as measured by reduction in glycosylated hemoglobin (HbA1c).
• To ascertain whether administration of LT4 for 16 weeks by titrating the serum TSH to 0.34 mIU/L - 1.7 mIU /L can improve serum lipid profile in Type II diabetic patients with NAFLD as measured by total serum cholesterol, high-density lipoprotein (HDL), low-density lipoprotein (LDL) and total triglycerides (TG).
• To ascertain whether administration of LT4 for 16 weeks by titrating the serum thyroid stimulating hormone (TSH) to 0.34 mIU/L - 1.7 mIU /L reduces the proportion of liver fat over body fat, which is reflected by fat in abdominal subcutaneous and visceral tissues, among patients with type II diabetes as measured by functional MRI on abdomen.

Subjects and Centres: A total of 50 eligible adult diabetic men with NAFLD will be recruited from 6 centres in Singapore - Changi General Hospital (CGH), Singapore General Hospital (SGH), Tan Tock Seng Hospital (TTSH), National University Health System (NUHS), Khoo Teck Puat Hospital (KTPH), Jurong Health (JH)

Inclusion and exclusion criteria are mentioned in the eligibility section.

Randomization: This is an open-labeled, single arm study.

Study protocol: After giving informed consent the subjects will undergo screening laboratory, abdominal imaging, ECG, review of medical history, medications for the past 6 months, and physical examination to determine eligibility. Once enrolled into the study the subjects will be treated with low dose LT4. There will be a titration phase of 12 weeks wherein the levothyroxine will be titrated to attain a TSH level between 0.34-1.70 mIU/L. It is possible that the desired TSH level may be achieved earlier than 12 weeks. It will not be considered protocol deviation if the 12-week titration period is not completed due to achieving the target TSH level earlier after following the defined titration method.

Subjects will undergo MRI/MRS at the beginning and end of study to assess changes in fatty liver content and changes in abdominal subcutaneous and visceral adipose tissue volumes. All study related liver MRI/MRS will be done at the Clinical Imaging Research Centre (CIRC) and will be read by a single imaging expert to minimize inter-reader variability.

Fasting glucose, HbA1c, insulin resistance, and lipid profile will also be ascertained before and after treatment. These tests will be done at respective institutions where the subject is recruited. The degree of insulin resistance will be determined using HOMA-IR which can be calculated easily online at www.hcvsociety.org/files/HOMACalc.html.

All subjects will still receive the standard care of treatment for NAFLD which is life style modification such as weight loss through diet and regular exercise. At each follow-up visit (visits 5 - 9), the investigator will solicit information from the subject on the latter's dietary and exercise habits and will counsel the subject accordingly on life style modification. BMI will be monitored during each follow-up visit. The random glucose and hbA1c tests will be done during each follow-up visit (visits 5-8) to monitor diabetes status. The subjects will also be encouraged to continue their diabetes follow-up.

Treatment: Titration of thyroxine is complicated and safety is paramount. Subjects will have their LT4 dose adjusted and monitored by highly-trained board-certified endocrinologists participating in this study. Investigators are provided with tables indicating the initial dosing and titration based on measured TSH values and timelines for subsequent TSH measurement. The investigators are authorized to exercise discretion to LT4 titration on case by case basis.

Study Duration: Maximum of 28 weeks: 12 weeks of pre-study drug titration and 16 weeks on low-dose LT4.

Statistical Analysis: Our primary endpoint is the absolute change in liver fat content from baseline after 16-week administration of LT4. Secondary endpoints are changes in HbA1c, total serum cholesterol, HDL, LDL and total triglycerides from baselines. The absolute change in liver fat content from baseline will be analyzed using one-sample two-sided t-test at a 5 % significance level. The same test will be applied to secondary endpoints.

For the primary endpoint and secondary endpoints, mean, standard deviation and 95 % confidence interval will be provided. Demographic and baseline data will be summarized with mean and standard deviation for continuous variables such as age, and count and proportion in each category for categorical variables such as gender and race. Adverse events and serious adverse events will be tabulated.

Inclusion Criteria:

1. Male between 21 to 60 years of age
2. Diagnosed with stable Type II diabetes mellitus (DM) with no changes in oral hypoglycaemic medications or dose for the last 2 months from the time of start LT4, and if on insulin < 10 units change in insulin dosage, documented by patient's medical records. The most recent HbA1C for the last 6 months from the time of start LT4 should be no more than 10%.
3. If the subject is on statin medication, there should be no change in the medication or dose of statin for the last 2 months from the time of start LT4
4. Baseline ALT <3 times upper limit of normal as per the institution's specified reference range , with a liver ultrasound showing presence of fatty liver (liver ultrasound will not be requested if a prior scan has been done within the past 6 months from the time of screening)
5. The IHL content on the MRI/MRS should be more than 10% to allow enrollment in the trial.
6. Baseline TSH levels between 1 - 10 mIU/L
7. Baseline heart rate <90 beats/min
8. Ability to provide informed consent

Exclusion Criteria:

1. Subject with history of viral hepatitis (except subject with history of viral A hepatitis or history of viral E hepatitis that was diagnosed at least 1 year before), hepatocellular carcinoma, liver cirrhosis, heart disease, osteoporosis, hyper/hypothyroidism, anxiety disorder, Graves' disease, thyroid/liver surgery, lactose intolerance, or malabsorption
2. Baseline estimated glomerular filtration rate (eGFR) < 60 ml/min
3. Currently on or within 6 months from the time of screening on either thyroxine, thiazolidinedione (TZD), oral T4/T3, anticoagulants (coumadin and warfarin), anti-viral drugs such as the protease inhibitors (ritonavir, indinavir, lopinavir), phenytoin, colestyramine, aluminium containing drugs (antacids, sucralfate), salicylates (> 100mg/day), dicumarol, furosemide, or sevelamer
4. Consumption of ethanol greater than 30g/day (i.e. 3 drinks/day or 21 drinks/week, with about 10g of alcohol per drink)
5. Has advanced liver disease with a baseline NAFLD fibrosis score of >0.675 (stage 3 or 4 fibrosis)
6. Has an implant or device in the body which is not safe for MRI scan
7. Baseline ECG findings considered to be clinically significant (e.g., ischemic changes, arrhythmias) by the Investigator(s)
8. Subject with history of claustrophobia
9. Baseline free T4 of more than the institution's specified reference range If a sole blood test result is deemed borderline according to the laboratory reference interval and not clinically significant, the investigator is authorized to exercise discretion.