IL-1 Signal Inhibition in Alcoholic Hepatitis

Histological improvement is defined as a reduction in lobular inflammation (regardless of cell type).

Improvement will be recorded as a binary Yes/No outcome. The outcome will be analysed as the proportion of patients, within each treatment group, with improvement on each individual component (polymorphonuclear cell infiltrate, ballooned hepatocytes and steatosis).

Where presence of bridging fibrosis or Cirrhosis at day 28 is the outcome. Firth logistic regression model was used. An intercept term and treatment indicator are the only predictor variables.

Number of patients in whom the histological degree of liver steatosis improved between baseline and day 28. Steatosis is assessed using an ordinal scale (1,2,3,4). There are four categories of steatosis grade: (1) <5%, (2) 5% to 33%, (3) >33% to 66%, and (4) >66% Ordinal logistic regression was used for statistical analysis

This outcome measure was included in the protocol as a secondary objective. However, no data was generated as HVPG is difficult to obtain and the outcome measure was therefore abandoned

Samples were not available. It was reported as a SAP deviation in the final statistical report.

MELD score is based on the following formula: MELD Score = (9.57 * ln(Creatinine)) + (3.78 * ln(Bilirubin)) + (11.2 * ln(INR)) + 6.43. Higher score is a worse outcome. Minimum value is around 9 and maximum value is around 40.

Predicts mortality in patients with alcoholic hepatitis by laboratory results and age. GAHS score is calculated after Forrest et al., 2007 where a score of 5 to 12 is assigned based on age, WCC, Urea, PT ratio or INR and bilirubin. Higher score means a worse outcome and a score greater than or equal to 9 is associated with a poor prognosis.

Outcome measure is the score at day 28 - score at day 0

mDF is calculated using the following formula: mDF = 4.6 x (Prothrombin time - control time) + Serum Bilirubin (μmol/l) / 17. A higher score is associated with poorer prognosis. A value more than 32 implies poor outcome.

Lille score is calculated as Exp(-R)/[1+exp(-R)] where R = [3.19 - (0.101*age in years)] + (1.47*albumin at baseline in g/dL) + [0.28215* (bilirubin at baseline - bilirubin at Day 8 in mg/dL)] - [0.206 * (if creatinine>=1.3 mg/dL at baseline)] - [0.11115*bilirubin baseline in mg/dL] - (0.0096*Prothrombin Time in seconds at baseline).

The Lille Model predicts mortality rates within 6 months. Scores >0.45 predict a 6-month survival of 25%. Scores <0.45 predict a 6-month survival of 85%.

Recommendations of the American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference; where SIRS represents the presence of 2 or more criteria out of following:

• Temperature < 36 ºC or > 38 ºC
• Heart rate > 90 beats/minute
• Respiratory rate > 20 breaths/minute or venous pCO2 <32 mmHg
• Leukocyte count > 12,000/mm3 or < 4,000/mm3 or band forms > 10%

Recommendations of the American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference; where SIRS represents the presence of 2 or more criteria out of following:

• Temperature < 36 ºC or > 38 ºC
• Heart rate > 90 beats/minute
• Respiratory rate > 20 breaths/minute or venous pCO2 <32 mmHg
• Leukocyte count > 12,000/mm3 or < 4,000/mm3 or band forms > 10%

Number of deaths per arm and hazard ratio.

The safety and tolerability of canakinumab will be assessed through the measurement of a number of participants with treatment-related adverse events.

Data were not available at time of final analysis.

The aim is to monitor baseline levels of cytokines linked to inflammasome activation and their changes after active IMP treatment. The proposed method and cytokines is as follows:

MSD 7-plex kit will be used for the detection of the following cytokines: IL-18, IL-1β*, IL-1ra, IL-6, IL-8, IFNγ** and TNF-α

ELLA 1-plex will be used for the detection of the following cytokines: IL-18Bpa

Improvement will be recorded as a binary Yes/No outcome. The outcome will be analysed as the proportion of patients, within each treatment group, with 95% confidence intervals.

Improvement will be recorded as a binary Yes/No outcome. The outcome will be analysed as the proportion of patients, within each treatment group, with 95% confidence intervals.

Where no/mild polymorphonuclear infiltration at day 28 is the outcome. A logistic regression model was used. An intercept term, baseline measurement and treatment indicator are the only predictor variables.

Where no megamitochondria at day 28 is the outcome. Firth logistic regression model was used. An intercept term, baseline measurement and treatment indicator are the only predictor variables.

Where there are 3 outcomes: (1) no or hepatocellular only, (2) ductular or canalicular, (3) canalicular or ductular plus hepatocellular. An ordinal logistic regression model was used. An intercept term, baseline measurement and treatment indicator are the only predictor variables.

There are four categories: (1) No foci, (2) < 2 foci per 200x field, (3) 2 to 4 foci per 200x field, and (4) > 4 foci per 200x field. Ordinal logistic regression was used.

There are three categories: (1) none, (2) few balloon cells, and (3) many cells/prominent ballooning. Ordinal logistic regression was used.