Editor's Note: This is part three of a three-part series on six men being hospitalized after taking part in TeGenero AG's Phase I trial. Part one and two ran Thursday and Friday.

Although co-stimulation now has been made notorious due to the Phase I trial that sent six previously healthy volunteers into intensive care within hours of receiving the experimental antibody, that antibody was not the first to target the co-stimulatory system.

One FDA-approved therapeutic targets the co-stimulatory system to treat rheumatoid arthritis, while several others are in clinical trials.

Given the similarities between activating and inhibiting molecules, it is easy to see why treatments that seem alike on the surface can have opposing effects in the clinic.

Co-stimulation has been a target both for attempts to rein in the immune system, for example in autoimmunity research, and for attempts to juice it up, in cancer therapeutics.

Orencia, which was approved in December for the treatment of rheumatoid arthritis, is the first molecule to receive FDA clearance to target the co-stimulation process and is an example of dialing down the immune system. It basically is soluble CTLA-4 and has the opposite effect of the anti-CTLA-4 antibody. Once administered, it competes with CD28 for binding to antigen-presenting cells and prevents T-cell activation.

If things go well, harnessing co-stimulation to strengthen the immune system in cancer treatments may not be far behind; Medarex Inc.'s MDX-010, an antibody to CTLA-4, is in Phase III trials to treat metastatic melanoma, and multiple Phase II trials for other cancers.

The antibody has the opposite effect of the soluble CTLA-4. It binds to CTLA-4 and prevents its binding to B7. That takes the brakes off of the immune system, leaving the field open for a more active immune response to tumor proteins.

Nils Lonberg, senior vice president and scientific director at Princeton, N.J.-based Medarex, was part of the preclinical development team for MDX-010. He said more than 600 patients have been treated with the antibody to date.

A report published in the Sept 1, 2005, issue of the Journal of Clinical Oncology detailed the responses of 56 patients that had been treated with two different dose schedules of MDX-010.

In the paper, the authors reported that the induction of autoimmune responses strongly correlated with antitumor activity of the antibody; 36 percent of patients who showed an autoimmune response responded to the drug, compared to a drug response rate of only 5 percent in patients with no autoimmune reactions.

The authors reported that in their study, all autoimmune reactions could be controlled with corticosteroids without interfering with the antitumor response, a finding they called "surprising and important." It joined an ever-growing list of surprising and important findings in the co-stimulation field.

CD28 also is being targeted in other clinical trials than the failed one by Parexel and TeGenero.

Carl June, a professor at the University of Pennsylvania and a co-stimulation expert, has conducted a clinical trial on activating the CD28 receptor to treat various blood cancers.

June's work, which he most recently reported on in the Feb. 15, 2006, issue of Blood, was ex vivo, meaning that the T cells were isolated from blood samples, stimulated outside the body and then re-introduced.

That technique has several advantages. For one thing, only a fraction of the T cells in a patient are activated with that method. For another, the activation is more controllable. "In layman's terms, we let them cool down," June told BioWorld Today. "We measured the cytokine release day by day."

That allowed the scientists to control how strongly the T cells were activated at the time point that they were reintroduced into patients. When activating the T cells within the body, as was the case with the TGN1412 trial, "whatever happens is going to happen," June said. "You can't dial it up or down."

June also pointed out one difference in trial logistics that would have turned out to be important to five of the volunteers if it had been followed in the TGN1412 trial: "When I did my trial, the FDA asked us to treat one patient at a time."

But the agency has no hard rule mandating sequential doses as opposed to simultaneous dosing, as its Office of New Drugs does not have guidance on the conduct of Phase I studies. The division has seen Phase I trials that have taken both approaches.

The FDA takes 30 days to review proposed Phase I protocols after an investigational new drug application has been submitted.

If a design passes muster and is reasonably safe to proceed, the study can begin. But if the agency determines otherwise, the FDA places the study on clinical hold, meaning it cannot be conducted lawfully in the U.S. until specific permission has been granted.

In addition, it seems unlikely that the FDA would make any sweeping changes in the near term to its case-by-case approach to reviewing clinical trial design.

A statement from the FDA's Office of New Drugs said it is "too early to speculate" on any changes to its current regulatory approach to Phase I studies, "since the investigation of what happened in the UK study is still ongoing and the results of that investigation will help to inform such decisions."

It is worth noting that the safety track record for Phase I studies conducted under FDA-approved INDs has "no incidents approaching the severity of the UK study having occurred in recent memory," the Office of New Drugs said.

Whatever setback the TGN trial has dealt co-stimulation targeting in the short term, June said he remains an optimist, noting that co-stimulation is hardly the first approach to have a rocky start in the clinic.

"Look at bone marrow transplant," he said. "All the initial patients died." But researchers and clinicians kept refining the techniques, "and now it's the standard of care."