Liver disease is one of the leading co-morbidities of human immunodeficiency virus (HIV) infection, and nonalcoholic fatty liver disease (NAFLD) is present in approximately 30-40% of patients with HIV infection. Nonalcoholic steatohepatitis (NASH) is a more severe form of NAFLD in which increased liver fat is also accompanied by inflammation, cellular damage, and fibrosis.

NAFLD is most prevalent in patients who also have increased visceral adiposity, and our group has previously shown that HIV-infected individuals with increased visceral adiposity generally have decreased growth hormone secretion. Tesamorelin is a growth hormone releasing hormone (GHRH) analogue that increases endogenous growth hormone secretion. Tesamorelin is FDA-approved for the reduction of visceral fat in HIV-infected individuals. In a previous study, treatment with tesamorelin in HIV-infected individuals selected for abdominal adiposity reduced liver fat. The current study is designed to test the effect of tesamorelin on liver fat and steatohepatitis in HIV-infected individuals who have NAFLD. The investigators hypothesize that tesamorelin will reduce liver fat and will also ameliorate the inflammation, fibrosis, and hepatocellular damage seen in conjunction with NASH.

Inclusion criteria:

• Men and women 18-70yo
• HIV-infection and treatment with a stable antiretroviral regimen for ≥ 6 months
• Hepatic steatosis as demonstrated by liver fat fraction ≥5% on 1H-MRS
• Hepatitis C antibody negative, or, if Hepatitis C antibody positive, either: a) known clinical disease, successful therapy ≥1 year prior to baseline and undetectable HCV RNA, or b) HCV resolved spontaneously and undetectable HCV RNA. Hepatitis B surface antigen negative at screen visit
• For females ≥50yo, negative mammogram within 1 year of baseline visit
• If use of Vitamin E ≥400 IU daily (in any formulation), stable dose for ≥6 months prior to study.

Exclusion criteria:

• Heavy alcohol use defined as consumption of more than 20g daily for women or more than 30g daily for men for at least 3 consecutive months over the past 5 years
• Use of insulin or thiazoledinediones (TZDs), or HbA1c ≥7%. Individuals with mild diabetes that is well-controlled with diet and/or oral anti-diabetic agents besides TZDs will be included. Use of oral anti-diabetics must have been stable for ≥6 months prior to study entry.
• Known diabetic retinopathy.
• Known cirrhosis, or Child-Pugh score ≥7, stage 4 fibrosis on biopsy, or clinical evidence of cirrhosis or portal hypertension on imaging or exam.
• Chronic corticosteroid use except intermittent use of topical steroid creams and/or prior short-term physiologic corticosteroid use in the ≤ 6 months prior to baseline visit
• Chronic use of methotrexate, amiodarone, or tamoxifen
• Known diagnosis of Alpha-1 antitrypsin deficiency, Wilson's disease, hemochromatosis, or autoimmune hepatitis
• Use of GH or GHRH within the past 1 year
• Change in lipid lowering or anti-hypertensive regimen within 3 months of screening
• HgB < 11.0 g/dL, CD4 < 100 th/mm3, or HIV viral load > 400 copies/mL
• Active malignancy
• For men, history of prostate cancer or evidence of prostate malignancy by PSA > 5 ng/mL
• Severe chronic illness judged by the investigator to present a contraindication to participation
• History of hypopituitarism, head irradiation or any other condition known to affect the GH axis
• Use of physiologic testosterone (men) or estrogen or progesterone (women) unless stable use for a year or more prior to study entry
• Routine MRI exclusion criteria such as the presence of a pacemaker or cerebral aneurysm clip
• Previous weight loss surgery
• For women, positive pregnancy test performed in a CLIA certified laboratory using a test with a sensitivity of at least 25mIU/mL, or breastfeeding.
• Known hypersensitivity to tesamorelin or mannitol
• Unwillingness to abstain from the conception process during the study (i.e., must agree not to participate in an active attempt to become pregnant or impregnate, donate sperm, or participate in in vitro fertilization)
• Unwillingness to use one (for males) or two (for females) reliable methods of contraception while engaging in heterosexual intercourse during the study. Acceptable methods for women include hormonal contraception (estrogen/progesterone or progesterone-only formulations) if stable for a year or more prior to study entry, intrauterine device, or barrier methods (condom, or diaphragm with spermicide). Acceptable methods for males include condom use. This requirement does not apply to women who have been post-menopausal for at least 24 consecutive months or have undergone surgical sterilization, or to men who have undergone surgical sterilization or have documented azoospermia.
• Not willing or able to adhere to dose schedules and required procedures per protocol

All randomized patients. One patient in the tesamorelin group discontinued before receiving study drug.

Adverse events reported herein are from the double-blind (tesamorelin vs. placebo) phase of the study which was 12 months in length as well as the subsequent open-label phase 6 months in length.

The primary analysis is from baseline to 12 months, during the "double-blind" phase. Adverse events from this phase are reported under the "Tesamorelin (Double-blind phase)" and "Placebo (Double-blind phase)" descriptors. 43 participants subsequently entered a 6 month open label phase during which all received tesamorelin. Data from this phase are not included in analysis, but Adverse Event data from the Open Label phase are included under the descriptor, "Open Label Tesamorelin Arm."