It is a while ago (2006) when TeGenero ran its ill-fated Phase 1 study of TGN-1412, a CD28 superagonist slated for studies in diseases like rheumatoid arthritis and B-cell CLL. Unexpectedly, in this FIM study, TGN1412 generated a cytokine release syndrome (CRS), some called it “cytokine storm”, a massive T-cell activation and cytokine release of unexpected proportions. All 6 exposed probands required urgent hospital care, many ended up in the ICU. Luckily, despite life-threatening complications in some probands, they survived the ordeal.
There were no easy explanations at first. Preclinical studies in-vitro and in animal models, including monkeys, did not give a warning of things to come. The testing models failed for reasons not understood at the time. Despite intense scrutiny from Regulators in Europe, there was no skeleton in the closet. In order to avoid similar disasters, EMEA issued a Guideline in 2007 in which new recommendations were made how to deal with biologicals that target immune functions [1].
The EMEA Guideline recommended careful individual dose administration and progression and warned against simultaneous cohort dosing. Had this procedure been followed by TeGenero, fewer cases would certainly have ended up in intensive care. More importantly, there was a recognition that the preclinical models were misleading because – for one reason or another – they did not mimic the human situation well enough. With no adverse events in the preclinical studies, TeGenero’s dose calculations based on NOAEL suggested a starting dose that proved to be way too high. The new Guideline recommended a different approach to FIM dosing: The MABEL approach relies on minimal anticipated effect size based on a variety of data sources including receptor occupancy calculations. It would have resulted in a much lower initial dose.
Interestingly, the dose response curve for TGN1412 is not a sigmoidal dose-response curve but bell-shaped. So, while a response is dose-dependent, doses either too low or too high have no effect at all. Not sure whether Paracelsus was aware of this in his days already.
Much progress has been made in our understanding of what went wrong and why there were no warnings. In a Nature Review article, Thomas Hünig, TeGenero’s lead immunologist and researcher, summarizes the lessons learned from TGN-1412 which should be required reading for those of us working on immunomodulatory drugs and biologicals [3] . According to the article, 3 new scientific discoveries provided explanations: (1) the interplay of TReg and CD4 effector memory T-cells was found to be different in humans compared to lab animals who had fewer effector/memory cells and thus were less prone to release cytokines; (2) the macaque model was not predictive because expression of the CD28 receptor changes over time with T-cell differentiation; and (3) while human PBMC do not react to TGN-1412 in standard activation assays, they do so when grown to high density in culture and with preconditioning.
The argument has been made that drugs targeting bacterial receptors are inherently safer than those which interfere with human cellular targets and immune activation pathways. This may be true but there are at 2 scenarios which still may create havoc: First, endotoxin release from bacterial disintegration may lead to toxemia, shock and MSOF, secondly, bacterial products could theoretically trigger a CRS as well..
The Schulz law is another variation on the dose-response theme: “For every substance, small doses stimulate, moderate doses inhibit, large doses kill”.
We have learned to live with some very toxic substances. Think about retinoids for acne, and thalidomide for multiple myeloma. Likewise, the learnings from TGN-1412 will certainly benefit the development of TAB08 (=TGN-1412 rechristened by TheraMAB) could still turn out to be a useful drug.
A 1000 times smaller dose was subsequently tested by TheraMAB in NHV without adverse events . An effect on regulatory T-cells was observed in a Phase 1 study [2]. Obviously, this was not a homeopathic dose by any standard.
Knowledge of the science is always incomplete when it comes to novel drugs. That’s why the post-mortem provided by Hünig and continued research by other researchers are so important [4] . Transitioning into man is a major step in drug development, and allometric scaling is not a safe platform in all instances.
Just remember: What’s safe for humans, may not be safe for your pet monkey (and vis-a-versa).
Abbreviations:
CRS cytotoxin release syndrome
FIM first-in-man
MSOF multisystem organ failure
PBMC peripheral blood mononuclear cells
References:
[1] EMEA Doc. Ref.EMEA/CHMP/SWP/28367/07
[2] O Dyer BMJ 2015;350:h1831
[3] T Hünig Nature Rev Immunol 12: 217, 2012
[4] R Stebbins Current Opinion in Biotechnology 2009, 20:673–677
