MICA: Can interferon gamma prevent infection in critically ill patients at highest risk? A phase II randomised controlled trial.

Lead Research Organisation: Newcastle University
Department Name: Translational and Clinical Res Institute


Patients admitted to an intensive care unit (ICU) are extremely vulnerable to developing new infections during their time in the ICU. These additional infections add to the risk of death, increase the time spent in the ICU and significantly increase healthcare costs. Antibiotics are given to too many patients in the ICU, partly because doctors are fearful that patients may develop new infections. Overuse of antibiotics leads to the emergence of "superbugs" that are resistant to antibiotics. For all these reasons, there is a pressing need for new methods to prevent infection in the ICU using treatments other than antibiotics.

Part of the heightened risk of infection in the ICU is because severe illnesses that cause patients to be critically ill in the first place "stun" the immune system, leaving immune cells less able to kill bugs. We know that a drug called interferon gamma (IFNg) can restore good function to patients' immune cells. The purpose of this proposal is to determine whether IFNg is safe, and whether it can reduce antibiotic use and infections, in patients admitted to the ICU who are at greatest risk of developing new infections.

We know that the risk of new infections is particularly high in two groups of patients in the ICU (a) patients who are unconscious and whose breathing is being done for them through a tube in their windpipe, which is connected to a ventilator machine, and (b) patients who require a machine to take over their kidney function and drugs to maintain a normal blood pressure. We also know that a particular blood test, called mHLA-DR, gives a good indication of whether the immune system has been stunned - a low mHLA-DR is associated with higher risk of developing infection and doing badly.

We have assembled a team across 11 ICUs in the UK. Patients in categories (a) or (b) in the paragraph above will be invited (through their relatives) to take part in the study. If consent is granted, a mHLA-DR blood test will be taken. If this is low, the patient will proceed into the study, which is a clinical trial. Patients will be allocated at random to one of four groups (neither the patient nor the research team will know which of the four treatments has been given). Each patient receives an injection under the skin (subcutaneously) once on 3 occasions over the next week. The groups are

1. Placebo on all 3 occasions
2. IFNg in the standard dose (100 microgrammes) on all 3 occasions
3. IFNg at half the standard dose (50 microgrammes) on all 3 occasions
4. IFNg where the first dose is 100 microgrammes, but the next 2 doses are either placebo or IFNg 100 microgrammes, depending on whether mHLA-DR tests are high or low, respectively.

In all study patients we shall record the antibiotics they receive, the time spent in ICU, whether they develop new infections, how many are still alive at 30 days, any potential side effects of IFNg, and mHLA-DR tests. Once 168 patients (42 per group) have entered the study, we shall pause the trial, and a statistician will analyse the results to see which of the IFNg groups has required the least antibiotics, while remaining safe. Effectively we "pick the winner" from the 3 IFNg groups and no more patients are admitted to the other two IFNg groups. We then resume the trial with only the placebo group and the "winner" IFNg group recruiting another 42 patients each. Once they have, we shall assess whether antibiotic use, death, new infections, length of stay in the ICU, and side effects are significantly less in the "winner" IFNg group than in the placebo group, and whether mHLA-DR goes up more in the "winner" IFNg group.

Technical Summary

Critically ill patients in intensive care units (ICUs) develop impaired immunity and are highly susceptible to ICU-acquired infections (ICU-AI), which increase mortality, morbidity and cost. Antibiotics are over-used in ICUs, contributing to antimicrobial resistance. There is a pressing unmet clinical need to reduce ICU-AI using non-antibiotic-based solutions. Interferon gamma (IFNg) improves immune dysfunction in critically ill patients. However, randomised controlled trials are lacking.

We propose a phase II, double-blind, randomised, placebo-controlled trial in 11 UK ICUs, in patients who are (a) intubated and mechanically ventilated, or requiring renal and blood pressure support, and (b) have low monocyte HLA-DR (mHLA-DR), denoting high risk for ICU-AI. The primary outcome is antibiotic-free days at 14 days (AFD-14).

The trial has a "pick the winner at interim analysis" design. 168 patients will be randomised 1:1:1:1 to 3 IFNg regimens or placebo, each given subcutaneously 3 times over one week. An interim analysis will identify the "winner" IFNg arm based on AFD-14, safety and change in mHLA-DR. The other 2 IFNg arms will be dropped. After interim analysis the trial will randomise 84 more patients 1:1 to the "winner" IFNg arm or placebo. Allowing for 10% drop out, 280 patients will be randomised in total.

The trial will determine if IFNg is safe and efficacious in patients at high risk of ICU-AI. The "pick the winner" design adds value by identifying the IFNg regimen most likely to demonstrate safety and efficacy, while describing each IFNg regimen's effect on mHLA-DR. The study has 80% power, alpha 5%, to detect a change in AFD-14 of 2. Secondary outcomes include 30-day all-cause mortality, infection-free survival, length of ICU stay, safety and change in mHLA-DR. Exploratory outcomes include health economic analyses and immune cell function.

If IFNg proves to be safe and efficacious, we are well-placed to move rapidly to a phase III trial.


10 25 50