In March 2006, six healthy volunteers in London received a new monoclonal antibody called TGN1412. Within hours, all six developed a severe cytokine storm and needed intensive care (Suntharalingam et al., 2006). All survived, but the trial changed how the industry and regulators approach first-in-human studies.
During my MSc in Drug Sciences at the University of Basel, the modules on Immunosafety and on Safety Assessment for First-in-Human Trials returned to this case again and again. This article summarises what went wrong and the practices that came out of it.
What TGN1412 was designed to do
TGN1412 was a "superagonist" antibody against CD28, a co-stimulatory receptor on T cells. Unlike normal antibodies, it could activate T cells without the usual second signal through the T-cell receptor.
The idea was to expand regulatory T cells and treat autoimmune diseases and certain leukaemias. Preclinical studies in cynomolgus monkeys had shown no serious toxicity at doses far above the planned human starting dose.
Why the animal data were misleading
The starting dose in humans was 0.1 mg/kg, roughly 500 times lower than the no-observed-adverse-effect level (NOAEL) in monkeys (Expert Scientific Group on Phase One Clinical Trials, 2006). On paper, this was a conservative margin.
The problem was biological, not mathematical. Later research showed that a key population of human T cells, the CD4+ effector memory cells, expresses CD28, while the equivalent cells in cynomolgus monkeys largely do not (Eastwood et al., 2010). Human T cells therefore responded far more strongly than the animal model predicted.
Standard in vitro tests had also failed to show the risk. Cytokine release was detected only later, when the antibody was presented to human cells in an immobilised, high-density format that better mimicked the situation in the body (Eastwood et al., 2010; Expert Scientific Group on Phase One Clinical Trials, 2006).
The lesson: from NOAEL to MABEL
The biggest change after TGN1412 was in how starting doses are set for high-risk products. Instead of relying mainly on the NOAEL from animal toxicology, developers now also calculate the MABEL: the minimal anticipated biological effect level.
MABEL is based on pharmacology, not toxicity. It uses data such as:
- receptor binding and occupancy at different concentrations;
- concentration–response curves in human cells in vitro;
- exposure–response data from relevant animal models;
- the potency of the molecule compared with similar products (Muller et al., 2009).
For agonists and immune-activating molecules, the starting dose is usually chosen from the MABEL or a pharmacologically active dose, whichever gives the safer margin. The European Medicines Agency (EMA, 2017) published guidance on first-in-human risk mitigation in 2007 and revised it in 2017.
Practical safeguards in today's first-in-human trials
TGN1412 also showed the danger of dosing several subjects at almost the same time. Today, trials of high-risk molecules commonly use (EMA, 2017):
- Sentinel dosing. One subject receives the active drug first and is observed before the rest of the cohort is dosed.
- Staggered administration. Subjects within a cohort are dosed with time between them, so an unexpected reaction can stop further dosing.
- Predefined stopping rules. Clear criteria for pausing or ending dose escalation are agreed before the trial starts.
- Facilities ready for emergencies. Units must be able to manage severe reactions immediately, including access to intensive care.
- Human-relevant in vitro assays. Cytokine release assays using human whole blood or peripheral blood mononuclear cells, in several formats, are now a routine part of the nonclinical package for immunomodulating biologics (EMA, 2017; International Council for Harmonisation [ICH], 2011).
Where my own work connects
My thesis investigated how pharmacological agents change cytokine and metabolic responses in activated macrophages. I quantified cytokines with ELISA and profiled cellular metabolism with FIA-MS.
That work gave me hands-on experience with exactly the kind of in vitro readouts that matter for immunosafety. It also showed me how strongly results depend on the experimental setup: cell type, activation state and the way a compound is presented all change the answer.
TGN1412 is the clearest reminder that an assay is only as good as its relevance to human biology.
Why this matters beyond the lab
The lessons of TGN1412 extend into clinical safety and pharmacovigilance. Safety teams monitoring trials of immunomodulators need to recognise early signs of cytokine release syndrome: fever, hypotension, rigors and rapidly rising inflammatory markers (Suntharalingam et al., 2006).
They also need to understand that serious unexpected events in a clinical trial trigger expedited reporting to regulators and ethics committees (ICH, 1994). Speed and clear documentation protect the next participant.
As a pharmacist, I have spent years identifying and documenting suspected adverse reactions in patients. Combined with my training in immunosafety and first-in-human safety assessment, that gives me a practical understanding of how safety is built into a product from the very first dose.
Key takeaways
- Animal models can miss human-specific immune biology, even with large safety margins.
- MABEL-based dosing, sentinel subjects and staggered cohorts are now standard for high-risk molecules.
- Human-relevant in vitro assays and clear stopping rules are central to first-in-human safety.
References
Eastwood, D., Findlay, L., Poole, S., Bird, C., Wadhwa, M., Moore, M., Burns, C., Thorpe, R., & Stebbings, R. (2010). Monoclonal antibody TGN1412 trial failure explained by species differences in CD28 expression on CD4+ effector memory T-cells. British Journal of Pharmacology, 161(3), 512–526. https://doi.org/10.1111/j.1476-5381.2010.00922.x
European Medicines Agency. (2017). Guideline on strategies to identify and mitigate risks for first-in-human and early clinical trials with investigational medicinal products (EMEA/CHMP/SWP/28367/07 Rev. 1). https://www.ema.europa.eu
Expert Scientific Group on Phase One Clinical Trials. (2006). Final report. The Stationery Office.
International Council for Harmonisation. (1994). Clinical safety data management: Definitions and standards for expedited reporting E2A. https://www.ich.org
International Council for Harmonisation. (2011). Preclinical safety evaluation of biotechnology-derived pharmaceuticals S6(R1). https://www.ich.org
Muller, P. Y., Milton, M., Lloyd, P., Sims, J., & Brennan, F. R. (2009). The minimum anticipated biological effect level (MABEL) for selection of first human dose in clinical trials with monoclonal antibodies. Current Opinion in Biotechnology, 20(6), 722–729. https://doi.org/10.1016/j.copbio.2009.10.013
Suntharalingam, G., Perry, M. R., Ward, S., Brett, S. J., Castello-Cortes, A., Brunner, M. D., & Panoskaltsis, N. (2006). Cytokine storm in a phase 1 trial of the anti-CD28 monoclonal antibody TGN1412. New England Journal of Medicine, 355(10), 1018–1028. https://doi.org/10.1056/NEJMoa063842
Member discussion