The return of the prodigal son and the extraordinary development route of antibody TGN1412 - lessons for drug development and clinical pharmacology

Innovative potential new medicines are a triumph of molecular design combined with advanced biological knowledge. The high tech products still often undergo painfully primitive assessments when they are finally tested in humans. TGN1412 was a chilling example of how such a sophisticated product could be destroyed by inadequate development.

The story of TGN1412 begins at the end of the last century when German researchers at the University of Würzburg developed an antibody that binds to rat CD28, a signalling molecule expressed on the surface of T cells. This antibody activated these cells independent of T cell receptor engagement [1]. Interestingly, the researchers speculated that the antibody preferentially activated the so-called ‘regulatory T cells’, and they further suggested that CD28-driven activation of regulatory T cells might be an effective treatment for autoimmune disease [2]. In 2000, with support from investors, the academic researchers founded the start-up biotech company TeGenero and generated antibody 5.11A1, a monoclonal mouse antibody that binds specifically to the human CD28 receptor and activates human T cells similar to the way in which the rat-specific antibody activates rat T cells [3]. To facilitate its use in clinical trials, the TeGenero scientists ‘humanized’ the 5.11A1 antibody to create the TGN1412 antibody.

Before antibody TGN1412 could be tested in humans, it was first tested in animals. The species chosen for the animal study was the cynomolgus monkey. This species was not chosen by accident. The CD28 receptor expressed in the cynomolgus monkey is highly homologous to the human receptor, and TGN1412 binds the CD28 receptors of both species with similar properties. The researchers therefore hypothesized that if TGN1412 produced no serious side effects in cynomolgus monkeys, the antibody could then be tested safely in humans. To err on the side of caution, the researchers included a safety margin for the first-in-man trial, using a starting dose that was only 0.2% of the maximum dose tested in cynomolgus monkeys.

TeGenero contacted Clinical Research Organizations (CROs) to discuss the pre-clinical data and the design and execution of a first-in-man trial. One of the key issues raised was the specificity of TGN1412. In particular, clinical pharmacologists questioned whether the antibody stimulates regulatory T cells only. Might it also stimulate other T cell populations that express CD28 receptors on their surface? What might happen if other T cell populations are activated as well? Is it possible to first test the antibody's selectivity in vitro using human cells prior to performing a first-in-man trial?

The specificity issue was by no means trivial, given that activated T cells produce cytokines, small proteins that play an essential role in regulating the immune response. Upon activation, different T cell populations produce different types of cytokines. For example, some cytokines activate the immune response and are therefore categorized as pro-inflammatory cytokines. On the other hand, anti-inflammatory cytokines have the opposite effect and down-regulate the immune response.

For their first-in-man trial, TeGenero contracted the London branch of the American CRO Parexel to conduct a trial using eight healthy male subjects. According to the study design, six men would receive the TGN1412 antibody and two subjects would receive a placebo. Parexel recruited eight healthy young men, explained that each participant would receive £2000, and obtained informed consent after explaining the study's aims and associated risks. The protocol had as primary end point just if the subjects tolerated the drug well enough. This is clearly low-tech evaluation of a high-tech product.

On Monday March 13, 2006, the TGN1412 antibody was administered to six healthy men at Parexel London's clinical research unit. Within an hour of receiving the antibody, the men complained of nausea, headache and severe back pain that rapidly increased in intensity. After a few hours, the health of the subjects deteriorated further, resulting in a loss of function of several vital organs, including the lungs and kidneys. As a result, they were transferred immediately to the intensive care unit of a nearby hospital. The two subjects in the study who received placebo remained healthy and were sent home.

At the intensive care unit, the condition of the six men who received TGN1412 deteriorated further. Participants’ partners and families were called and asked to come to the hospital, where they received the news that their loved ones were gravely ill and potentially facing a fatal outcome.

Because TGN1412 was a first-in-class antibody that had not been administered to humans before and because the doctors had no experience dealing with drugs with a similar mechanism of action, it was not clear how the men should be treated. Through media channels, experts were consulted for advising the doctors regarding treatment. Fortunately, the doctors saved the patients’ lives. However, the subjects suffered permanent damage, fingertips and toes became necrotic and had to be amputated. In July 2006, TeGenero was bankrupt, and following a legal battle, the men received compensation for their injuries. With the dramatic outcome of the trial and the bankruptcy of TeGenero, the TGN1412 antibody seemed to have no future as a therapeutic drug.

In September 2006, the British doctors published their findings regarding the treatment of the critically ill subjects [4]. They reported that the six subjects who received TGN1412 rapidly experienced a steep increase in the production of pro-inflammatory cytokines, a phenomenon known as a cytokine storm. In November 2006, the Expert Scientific Group, which was established by the Secretary of State for Health to investigate the TGN1412 trial, concluded that all regulatory requirements (specifically, good manufacturing practice and good clinical practice guidelines) were satisfied by TeGenero and Parexel [5]. Nevertheless, TeGenero was heavily criticized by independent researchers [6,7]. In particular, critics argued that TeGenero's preliminary investigation was incomplete and the full set of preclinical data was not presented to the reviewing regulatory authorities [8]. In addition, previous clinical trials using antibodies with a similar, albeit not precisely the same, mechanism of action had already shown that antibodies that activate T cells can lead to a cytokine storm.

The starting dose of TGN1412 used in the clinical trial was also a source of intense criticism. Calculations revealed that the starting dose was sufficient to cause approximately 90% of all CD28 receptors in the human body to be occupied by TGN1412 antibodies [5]. This high level of receptor occupancy can lead to the robust activation of multiple T cell subpopulations and thus drive extremely high level production of various cytokines. Among pharmaceutical companies, for biotherapeutics with potentially agonist modes of action on key body systems, a starting dose is selected that corresponds to a maximum receptor occupancy of 10% [9]. Strikingly, the researchers had failed to include such a calculation when designing their clinical protocol.

Despite the devastating outcome of the London first-in-man trial and despite criticism from the scientific community, a Russian investor purchased the rights to antibody TGN1412 at the end of 2006. The antibody was subsequently renamed TAB08 and was developed further by the young Russian biotech company TheraMAB. Meanwhile, researchers at the University of Würzburg continued to investigate the London trial in an attempt to determine the cause of the severe outcome. They developed an in vitro laboratory test that enabled researchers to study closely the effect of TGN1412/TAB08 on human T cells [10]. This test, which was not available at the time of the London study, yielded important information that could be used for a rational approach to determine a safe starting dose.

In addition, British researchers discovered why TGN1412 caused serious side effects in humans but not in cynomolgus monkeys. In cynomolgus monkeys, T cells that develop into so-called CD4+ effector memory T cells lose their CD28 receptors [11]. Thus, these cells would not be activated by the TGN1412 antibody. In humans, however, CD4+ effector memory T cells retain CD28 receptors on their surface. Thus, human CD4+ effector memory T cells are still activated by TGN1412 and therefore rapidly produce pro-inflammatory cytokines.

Armed with the results of the new laboratory tests and new insights into the mechanism by which TGN1412/TAB08 activates human T cells, Russian researchers at TheraMAB and doctors at the Hospital for Emergency Medical Care in Yaroslavl, Russia were now in a position to re-test the antibody in healthy volunteers. For their study, the researchers chose an extremely low starting dose. Their dose was 0.1% of the dose used in the London study (and therefore 0.0002% of the maximum dose used in the original cynomolgus monkey study). Calculations showed that at the new starting dose, only 1% of human CD28 receptors would be bound by TGN1412/TAB08 antibodies. During the clinical study, the dose was increased by small increments in different cohorts, and all subjects were monitored closely. No serious side effects were observed, and the results of the new trial were published in April of last year in the European Journal of Immunology. The results indicated that 5% of the dose used in the London trial selectively activated regulatory T cells, but not pro-inflammatory T cells [12].

The Russian researchers are conducting a follow-up study in which TGN1412/TAB08 is being administered to patients with rheumatoid arthritis [ClinicalTrials.Gov registration number: NCT01990157] [13]. The company's ultimate goal is to develop the antibody into a drug that will down-regulate the inflammatory response in rheumatoid arthritis by selectively activating regulatory T cells. So 8 years after its dramatic and troubled start, the TGN1412 antibody is back in the clinic and may be developed into an innovative drug for treating autoimmune diseases.

The extraordinary drug development route of TGN1412 demonstrates that innovative and potentially risky drugs can be tested safely in humans if researchers have sufficiently detailed insight into the mechanism of action and if informative preclinical studies and biomarkers are available for accurately predicting the effect in humans, thus providing a rational approach for determining a safe starting dose [14]. With the potentially triumphant return of the TGN1412 antibody, scientists and regulators can be motivated to reduce their focus on regulations and increase their focus on biological science. Developing a new medicine in humans requires pharmacological insight, choice or the development of methodology for the evaluation of the effect and knowledge of the clinical conditions. TGN1412 was turned from a disaster into the potentially useful medicine it already was through good clinical pharmacology and dose selection based upon pharmacological principles. In the BJCP, as we stated before, we will keep our focus on studies of sophisticated new medicines studied with appropriately sophisticated methodology [15].