Cholic Acid for Hepatic Steatosis in Patients with Lipodystrophy: A Randomized, Controlled Trial

Setting and Participants

All patients were evaluated at the Clinical and Translational Research Center (CTRC) at UT Southwestern Medical Center, Dallas, Texas. A written informed consent was obtained from all participants, and the study was approved by the Institutional Review Board of UT Southwestern. Thirty-seven patients with a clinical diagnosis of genetic or acquired autoimmune lipodystrophy (not HIV-associated) underwent screening with ¹H magnetic resonance spectroscopy (MRS) for liver fat. Ten subjects did not qualify after screening (8 had low levels of liver fat, one was too obese for the MRI machine, one had a history of alcohol use), 8 subjects lost interest, and one could not be contacted (Figure 1).

Inclusion criteria for the trial were as follows: patients with lipodystrophy as diagnosed by clinical criteria, hepatic steatosis (>5.6% hepatic TG content) as demonstrated by ¹H MRS, age 10-70 year, and alcohol intake of less than 40 g per week. Patients with the following were excluded: other liver diseases such as chronic viral hepatitis, autoimmune hepatitis, primary biliary cirrhosis, biliary obstruction, Wilson's disease, hemochromatosis or α-1-antitrypsin deficiency; treatment with drugs associated with steatohepatitis: corticosteroids, high dose estrogens, methotrexate, amiodarone, calcium channel blockers, sulfasalazine, naproxen, or oxacillin in the 6 months prior to the study; decompensated liver disease; hepatocellular carcinoma; congestive heart failure, cerebrovascular disease, respiratory failure, renal failure (serum creatinine >2 mg/dL), acute pancreatitis, organ transplantation, serious psychiatric disease, and malignancy; acute medical illnesses; known HIV infection; current substance abuse; pregnant or lactating women; hematocrit of less than 30%; weight loss during past 3 months; and hypersensitivity or intolerance to CA or any components of its formulation. Additional exclusion criteria included the use of drugs which can potentially decrease hepatic steatosis: ursodeoxycholic acid, high-dose vitamin E, betaine, acetylcysteine and choline. Thiazolidinediones were allowed if the dose had been stable for three months prior to screening.

Design Overview, Randomization and Interventions

This study was a randomized, double-blind, placebo-controlled, cross-over trial (Figure 1). Patients underwent a screening evaluation to determine their eligibility for the trial. For those who were found to be eligible, they continued their usual diet and other lifestyle measures without changing any medications for 1 month in order to establish a baseline state.

All patients were studied during three hospitalizations, each lasting for three days, and two outpatient visits (Figure 1). One month after screening, patients were hospitalized to the CTRC for a period of three days, i.e., the baseline visit. Fasting blood samples were obtained on 3 consecutive days during the inpatient evaluation, and the average of the three measurements was used for data analysis. All patients underwent ¹H MRS to determine liver fat during this hospitalization. At 6 and 12 months, patients were again admitted for 3 consecutive days; during these visits, all the studies mentioned in the baseline visit were repeated. The patients returned at 3 and 9 months for fasting blood samples and liver ¹H MRS during outpatient visits.

Following the baseline hospitalization, the patients received CA or an identical placebo in the dose of 15 mg/kg/day (divided twice daily) for a period of 6 months and then received the other treatment (CA or placebo) for the next 6 months. The maximum dose of CA was 1500 mg/day. CA and matched placebo were supplied by Global Strategic Connections, LLC (Troy, MI) in 250 mg capsules.

At each study visit, an inquiry was made about any side effects of medication, in particular, gastrointestinal side effects. Medication compliance was assessed by counting the remaining number of pills at the completion of each study period. If a patient was unable to tolerate the full dose, the dose could be reduced by the investigators. Patients were asked to keep doses of their other medications constant for the duration of the study.

Outcomes

The ¹H-MRS studies were performed at least four hours post-prandially with the patients lying prone as described previously ²⁴. Briefly, image-guided, proton-localized, MRS and high resolution T1-weighted imaging were performed on a 1.5 T Gyroscan Intera whole body system (Philips Medical Systems, Best, The Netherlands) with the following parameters: repetition time (TR) of 3 s, spin echo time (TE) of 25 ms, and 1024 data points over 1000 kHz spectral width. Volume of interest (voxel 30 mm³) was selected in the upper right hepatic lobe, taking care to avoid vascular structures. Spectra were processed and resonances quantified using a standard analysis package (NUTS; ACORNNMR, Fremont, CA). The hepatic TG content was expressed as the ratio of fat and water resonance peaks.

Blood was obtained after overnight (12 h) fast daily for chemistry profile (SMA-25) and lipoprotein levels while the patients were admitted. Serum chemistry profile including the aminotransferases, lipoproteins, gamma-glutamyl transpeptidase (GGT), electrolytes etc. were measured by autoanalyzer (Quest Diagnostics, Irving, Texas). Glycosylated hemoglobin (HbA1c) levels were measured by an immunoturbidimetric colorimetric assay (Quest Diagnostics, Irving, TX).

Primary and Secondary Endpoint Variables

The primary end-point variable was reduction in the liver TG content as measured by ¹H MRS. Reductions in the levels of serum alanine and aspartate aminotransferases (ALT, AST), GGT and TG were secondary end-points of the study.

Statistical Analysis

We determined that a sample of 12 patients completing the 12 month crossover study would provide a power of 0.80 at an alpha of 0.05 to detect an absolute treatment difference in hepatic fat content of 3.5% with the standard deviation of the difference of 4.0%. We therefore planned to enroll at least 16 subjects assuming a maximum 30% rate of loss to follow-up.

A block randomization with block size of two was generated using SAS version 9.2 version 9.2 (SAS Institute, Cary, NC, USA). At the beginning of the study, a randomized treatment sequence was prepared by the study statistician who sent it directly to the investigative pharmacy. The pharmacist assigned participants to the interventions according to the treatment sequence. Patients were recruited by the study personnel. All participants and study personnel and those evaluating outcomes (e.g., ¹H-MRS results) were blinded. Only the pharmacist was aware of the intervention sequence. The code was broken upon study completion.

All data on patients who had at least one 3 month measurement of hepatic TG content during the placebo or CA phase were included in the analysis (n=15). For continuous outcome variables, statistical analysis of this two-phase crossover study was performed using mixed linear model repeated measures analysis with the study subject modeled as a random effect. The mixed-effects analysis was also used to assess a possible treatment sequence effect. All hypothesis tests were two-sided and a p-value < 0.05 was considered statistically significant. Results are reported as median and interquartile range. Statistical analyses were conducted using SAS.

Patients

Recruitment began in December 2006, and the last patient finished the 12 month follow-up visit in April 2011. Eighteen patients were enrolled in the study and 12 of them completed both the CA and Placebo phases, whereas an additional 3 patients had 3 month data (Figure 2). The five dropouts included three who were unable to be contacted, two who had hospitalizations and medication changes for unrelated illnesses, and one who had an implantable cardioverter-defibrillator placed for pre-existing cardiomyopathy due to which MRS could not be performed.

Table 1 describes the baseline characteristics of all patients enrolled in study. At screening, the mean (SD) hepatic TG content was 23.9% (13.4%), and the range was 6.1%-60.3%. Serum ALT concentration ranged from 11 to 168 U/L; AST ranged from 12 to 55 U/L; GGT ranged from 16 to 72 U/L; and serum TG concentration ranged from 155 to 3455 mg/dL. Twelve patients had a previous diagnosis of diabetes mellitus.

Nine patients had familial partial lipodystrophy of the Dunnigan variety (FPLD) due to mutations in the lamin A/C (LMNA) gene. Eight patients had other types of familial partial lipodystrophy. One patient had acquired generalized lipodystrophy.

Primary and Secondary End Points

The sequence in which the patients received CA or placebo had no effect on the results. Compared with placebo, CA therapy did not result in any significant reduction in hepatic TG concentration (geometric mean difference, 13.8%; 95% confidence interval [CI] −19.7% to 61.1%). Similarly, there was no significant change in serum TG levels (11.4%; CI −17.7% to 50.8%). Individual responses for hepatic TG concentration and serum TG levels are shown in Figure 3 and overall results are reported in Table 2. CA therapy did not lower ALT (−3.0%; CI −24.3% to 24.2%), AST (2.5%; CI −17.6% to 27.7%), or GGT (1.6%; CI −18.0% to 26.0%) levels. Mean difference in HbA1C was 0.41% (CI −0.23% to 1.05%); however, it was not statistically significant (p = 0.18). Absolute difference in weight was −0.23 kg (CI, −2.02 to 1.56 kg) for CA relative to placebo; however, it was also not statistically significant.

Adverse Events

CA was well tolerated except for two patients who developed diarrhea and excessive flatus while taking CA. Their symptoms resolved after reducing the dosage of CA to 5-10 mg/kg/day. Three other patients reported mild diarrhea which subsided within one to two weeks without any dose adjustment.

Four patients had adverse events during the trial. One patient developed gastroenteritis requiring intravenous fluids. Two patients developed flu-like symptoms with vomiting and diarrhea and also required intravenous fluids for volume depletion. One patient with multiple co-morbidities had four hospitalizations for fluid accumulation due to heart failure, chronic obstructive pulmonary disease, and pneumonia. All of these adverse events occurred while the patients were taking placebo.