Primary Prevention Hepatocellular Carcinoma by Metformin

Hepatocellular carcinoma (HCC) is currently the first cause of death of patients with compensated HCV cirrhosis.Despite progresses,existing therapies are limited in their ability to prevent recurrences. Even diagnosed at early stage, long-term prognosis remains poor due to the high rate of recurrence after local treatments. Liver transplantation the only long-term curative treatment is limited by advanced age, comorbidities or the shortage of the graft It concerns less than 5 % of HCC patients . Therefore, the best approach to reduce mortality remains the reduction of HCC incidence.

Abundant observational studies have related a relation between insulinoresistance occurrence and outcome of many cancers. The level of IR assessed by the HOMA index have been recognized as an independent predictive factor of HCC occurrence in patients with compensated viral C cirrhosis. Metformin, a Type 2 diabetic treatment drug, inhibits hepatic gluconeogenesis and increases the stimulation of the glucose uptake in muscle.

Independently of its' anti diabetic effects, Metformin is credited of anti tumoral, anti oxidant, anti inflammatory, and anti angiogenic properties.

Amount epidemiological and experimental data have demonstrated the anti tumoral and chemopreventive effect of metformin in certain cancers.

From our cohort of patients with compensated HCV cirrhosis and not treated by insulin, we have observed that the level of IR assessed by the HOMA was a strong and independent risk factor of HCC occurrence and liver related death. We have also observed in our cohort of diabetic patients with compensated HCV cirrhosis, that treatment by Metformin was associated with a decreased risk of HCC occurrence and liver related death.

HYPOTHESIS

Treatment with metformin could decreased the HCC occurrence and liver related death or transplantation.

MAIN OBJECTIVE

Evaluation the impact of Metformin treatment on HCC occurrence and liver related death in patients with compensated HCV cirrhosis and Insulinoresistance SECONDARY OBJECTIVE

• Occurrence of decompensation of the cirrhosis (ascite, sepsis, encephalopathy, haemorrhage)
• Evaluation of the treatment tolerance

MAIN CRITERION JUDGMENT

Rate of HCC occurrence or liver related-death or transplantation.

SECONDARY CRITERION JUDGMENT

• Occurrence of decompensation of the cirrhosis (ascite, sepsis, encephalopathy, haemorrhage),
• Tolerance

STUDY ASSESSMENTS

The patient of CIRVIR cohort meeting the inclusion criteria will be invited to participate to this study.

During their next visit, the hepatologist, will give full verbal and written information regarding the objective procedures of the study and the possible benefice and side effects of the treatment. A write informed consent will be obtained from all patients who agree to participate to the study.

The treatment period will begin following randomization. On day M0 baseline measurements will be taken and recorded, and metformin administration will be begun. In order to optimize the treatment tolerance, it will be suggested to the patients to take the pill during or at the end of the lunch. During the first week, the posology of the placebo and metformin will be 500 mg at the breakfast. After, the posology will be increased every week as follow: 500 mg morning and afternoon, then 1000 mg morning and afternoon (2000 mg per day). In case of intolerance, the maximum posology tolerated will be maintained. In fact regarding the primary data of the trial regarding the effect of metformin on colonic polyp, it seems possible that low dose of metformin are potentially active This treatment will continue until the end of the study.

FOLLOW UP

Patients will be seen at one month and followed every 3 months. Clinical evaluation and HCC screening are planed In CIRVIR cohort study, Every 6 months.

Duration of Treatment per patient:

• 36 months

Duration of Trial Recruitment:

• 24 months

PARTICIPATING CENTERS : 26

NUMBER OF SUBJECT

In order to demonstrate a reduction of 40% (HR 0.6) of events under metformin vs placebo with 80% power and 5% two-sided alpha risk, 200 patients per arm are necessary.A sample size reassessment will be made after 50% and 75% of patients included based on predictive power calculation.

We estimated that 5% of patients will not tolerate the treatment in the first month, and that 5% more will be lost to follow or not compliant to treatment during the follow up period. Therefore, the number of patients to be included is 222 patients per group.

STATISTICAL ANALYSIS

Clinical data of all the patients will be prospectively collected in a computerized database

Populations analyzed The main analysis will be based on the intent-to-treat population (ITT) of all randomized patients

In addition an explanatory analysis (PP) of all patients randomized \& treated without major protocol violations/deviations will be carried out. Pre-defined major protocol violations/deviations are:

1. missing data for the primary efficacy endpoints
2. no study drug received
3. violation of inclusion criteria
4. Additional protocol violations will be possibly defined during the blind data review

Statistical tests. Main criterion: rate of HCC occurrence and liver related-death or transplantation.

The cumulative incidence of HCC and liver-related death or transplantation will be compared according to metformin treatment at inclusion using the log-rank test.

In addition, univariate Cox regression models will be used to identify predictive factors of primary endpoint.

For each endpoint, variables with a P value less than 0.10 in the univariate analysis predicting outcomes will be entered into stepwise Cox regression multivariate models. For sensitivity analyses, the incidence of HCC will be also adjusted on usual risk factors. The same models considering competing risks will be tested using the Fine and Gray test.

Secondary criteria : Occurrence of decompensation of the cirrhosis (ascite, sepsis, encephalopathy, haemorrhage).

Comparisons between groups will be performed first in a univariate manner using the χ2 test or the Fisher-exact tests. Multiple logistic regression models will be used to assessed a possible difference between groups when adjusted on parameters known or identified during the study as possibly affecting these outcomes.