Drug trials - what went wrong.

The clinical trials disaster in London has intimidated people from taking part in future clinical trials. A recent poll in AOL indicated that 80% would not take part in a drug trial. It is important to look at what happened.

The drug being tested was a monoclonal antibody directed at a molecule on the surface of T cells called CD28. It is one of the co-stimulatory molecules on T cells. Normally in order to stimulate T cells you need to stimulate both the T-cell receptor (TCR) complex (usually done with anti-CD3 in the test tube) and a co-stimulatory molecule. However, there is a subclass of anti-CD28 antibodies known as superagonists which can stimulate T cells without involving the TCR. They do this by binding to a different site on the CD28 molecule, so that adjacent molecules are lined up in a linear array on the cell surface, which facilitates signaling.

An important property of CD28 stimulation is that it is crucial for the survival and proliferation of regulatory T cells. Regulatory T cells have suddenly assumed a great importance in immunology. They are recognised by being positive for CD4 and CD25 and have a main function in preventing the growth of autoreactive T cells. Mice that lack regulatory T cells develop autoimmune disease. Administration of superagonistic anti-CD28 to mice with autoimmune diseases produces remissions. There is also some suggestion that these T cells have a role in controlling the growth of lymphoid tumors.

In CLL patients with autoimmune disease like AIHA and ITP the regulatory T cells are suppressed. Not only did this antibody appear promising in autoimmune diseases like rheumatoid arthritis and diabetes, but in CLL it might controle AIHA and ITP and possibly the CLL itself.

So what went wrong? One of the problems of stimulating T cells is that they release cytokines. Among these is interleukin-2, which is licensed for the treatment of melanoma and kidney cancer. It is extremely toxic. It cause a high fever, shivering attacks, diarrhea, low blood pressure and makes the patient feel dreadful. So-called capilliary-leak syndrome develops. Fluid leaks from the capilliaries of the lungs, which become waterlogged, so that breathing is difficult. The blood pressure falls so that the kidneys shut down and renal failure develops. Other effects include damage to the liver and bone marrow. The activated lymphocytes visit the heart, so that sometimes heart attacks occur. Interleukin-2 induces the relase of other cytokines - the so-called cytokine storm. To me this sounds like what has happened to these volunteers. Interleukin-2 stays in the blood for 7 minutes. Humanized monoclonal antibodies hangs around for an average of three weeks.

When the anti-CD28 was given to mice there was no evidence of cytokine storm. Mouse CD28 is very similar to human CD28, and the human anti-CD28 cross-reacts with the mouse molecule. As far as I can tell it had been safe also in primates, probably macaques. Nevertheless, in the trial, just to be safe the volunteers were injected with one five hundredth the dose of what was safe in the mouse. No-one was expecting it to be anything other than safe in the volunteers.

Is there any evidence that the trial was improperly conducted? This was a phase I trial intended to test the safety of the drug and to determine a safe dose. Often cancer drugs are tested on patients who are terminally ill who have a slight chance of benefitting from the drug; here a decision was made to use healthy volunteers. In retrospect, using sick patients might have resulted in them all dying of the side effects. The drug was given to eight volunteers with a new volunteer being started every few minutes. This seems foolhardy, but is a common practice. Obviously, had they been injected on successive days the trial would have stopped after one volunteer. But with the background of a safe animal trial behind them, no-one was expecting anything untoward. The volunteers were paid £2000 each. It is not surprising that eight healthy young men felt the risk worth taking. One young man testified in the newspapers that he had paid off his mortgage by taking part in phase I trials.

Does this invalidate the use of animal trials? There is no point in protesting that animals are identical to humans. If that were so there would never be any need to test on humans after a drug has been shown to be safe in animals. Clearly there are diferences. But hundreds of drugs have never been tested on humans because they were too toxic to animals. If the animal tests were omitted there would be many more like the London Six.

My conclusion is that this was an idea with great potential benefit. So far the only criticism that I can make was that they should have staggered the doses so that the side effects would have affected fewer volunteers, but hindsight is fine thing.