Skip to main content

Clinical Trials

Tegenero, In Detail

The New England Journal of Medicine has published an authoritative wrap-up of the Tegenero/TGN1412 case. This, you’ll remember, was the T-cell stimulating antibody trial that went disastrously wrong, sending six first-in-man voluteers into intensive care. (They remained there for one to three weeks, but all of them survived). As minor side effects, this event also sent the company into bankruptcy and the drug candidate straight down the waste chute.
The article makes for grim reading, and I’d be interested to hear what some of the med-bloggers have to say about it. I’m no MD, but the patient charts on admission to the ICU look pretty terrifying to me – pulmonary and renal failure, coagulation throughout the vasculature, severe (and surprising) near-total loss of lymphocytes and monocytes, and much, much more. The phrase “empirical treatment” shows up a lot in the account of their cases, which I take to mean “what seemed reasonable, since we’d never seen anything quite like this before”.
As it turned out, the empirical treatment – intubation, dialysis, transfusions, whacking doses of steroids and anti-IL-2 receptor antibodies – seems to have done the trick. This was a “cytokine storm”, a known immune phenomenon never observed in such isolation before. (It’s usually set off by infection or some endotoxin).
An accompanying perspective article talks about some the issues that were tossed around on this site at the time (a href=””>here – scroll back to March), such as the similarities (and differences) between the CD28-targeting Tegenaro antibody and the comparatively successful ones targeting CLT-4. No one is still quite sure why TGN1412 did what it did, but the authors have a couple of suggestions: for one thing, the affinity of the antibody was probably quite different in humans than in the other species used preclinically. In primate studies, the animal were dosed with the exact same antibody (anti-human-CD28) used in the clinical trial, but it surely has weaker binding to the primate T-cell receptor.
Another factor, which as they point out is not often appreciated, is that laboratory animals (particularly rodents) have much more naïve immune systems than wild-type animals (like us). The clinical trial subjects surely had far more memory T cells, activated by previous exposures to all sorts of antigens, than the mice used in the early models. Perhaps this added to the trouble.
At any rate, we don’t seem to be hearing much about how TGN1412 might still go back into the clinic, like we were at the time. The NEJM authors correctly point out that before anyone goes after any of the costimulatory T-cell receptors again, we’re going to need to know a lot more than we do now. And even then, you have to think, it’s going to require an awful lot of nerve.

12 comments on “Tegenero, In Detail”

  1. Robin Goodfellow says:

    Were any studies done in primates with species specific anti-CD28 (e.g. anti-monkey-cd28 in monkeys)?

  2. Petros says:

    At least one of the treated volunteers has recently been reported to be dispalying cellular changes suggesting that he will develop cancer!

  3. Luke, I am your fungus! says:

    maybe army could find some use for the stuff…

  4. GATC says:

    Hey Luke (the fungus man), chillingly so……see this link for a potentially infectious version of “Night-of-the-Cytokine-Killer”
    And yet the NIAID and others are still farting around with contracts for the mundane like anthrax vaccines and rapid detection systems. Wait until these recombinant versions start hitting the streets…………

  5. financeguy says:

    Regarding Sanofi and plavix: If Sanofi has marketing rights outside US, why does the US patent ruling matter to them? Wouldn’t European patents still protect Sanofi’s right to sell Plavix in Europe?

  6. Still Scared of Dinosaurs says:

    NIAID isn’t “farting around” with anthrax vaccine because we need better protection from the disease. It’s doing it because the current vaccine has so many unpleasant side effects.
    And I’m sure many others can set me straight the following, but the “human/non-human” issue with an antibody like this seems to me to have at least 4 components.
    1) Is the target receptor upon which the ab is based the human receptor?
    2) Is the system from which the end of the ab is derived human? Seems to me you often get mouse antibodies targeting human proteins as the starting point.
    3) Is the tail end of the ab a human antibody and what is it likely to be doing? Different tails have different affinities for their receptors and can send different signals. If the tail is adapted to signal NK cells to kill the cell to which the ab is attached, you can stimulate it all you want but it may not live that long.
    4) Has the entire molecule, to the extent possible, had non-human structures replaced with human ones. I believe this is what they mean by “fully humanized”. The more foreign a molecule, the more likely you are to have neutralizing antibodies develop (anti-antibody antibodies…this too I gotta deal with?)
    I’m no expert on this stuff, and most importantly I have no idea how relevant each of the above may be to what happened in this case. I’m pointing this out partly to reinforce the idea that it’s better to think of antibodies as battleships rather than bullets.

  7. Paul says:

    Your comments on antibody affinities are interesting. What do you think the problem was here — that the antibody was too specific (for want of better terminology — overaffinity or overbinding)? Does that then say something about the target or about the antibody technology (don’t know if it was a human or humanized and whose technology it was)? Could a less specific therapy be better than a more specific one, especially since our knowledge on the reality of specificity is delusional at best.

  8. Tom says:

    See recent issue of:
    Science 21 July 2006:
    Vol. 313. no. 5785, pp. 308 – 309
    Short summary – Most mAb drugs work by binding to cells to flag them for killing or to shut down a receptor. This mAb is an agonist to CD28, preferentially binds T-regs (Regulatory T-cells), intended result of treatment is to selectively activate T-regs which would slow down immune system for autoimmune diseases.
    But – CD28 present on all T-cells, other T-cells outnumber T-regs by more than 10:1, mAb activates 1st T-cell it comes to. mAb doesn’t supress immune system as intended, but activates instead. All subjects likely to die of cancer in a few years.

  9. mordina says:

    The monoclonal antibody TGN1412 appeared to unleash a cytokine storm which resulted in the near fatal response in all 6 patients (2006) injected simultaneously in the UK. Unfortunately the trial was launched in humans for the first time. It had been noted that injections of other non-human vertebrates had no untoward effects.
    Recently 2007, in the US another company, Targeted Genetics Corp, reported a death in their trial of tgAAC94 for the treatment of inflammatory arthritis. This product, using a recombinant AAV vector to deliver a DNA sequence that encodes a soluble form of the TNF-alpha receptor, was injected for the 2nd time directly into the affected joint of a subject with inflammatory arthritis while the patient was taking concurrently 3 other potent arthritis drugs. The 1st injection was asymptomatic. It appears that this last injection provoked a cytokine storm. Again there were breaches of clinical research standards, ineffective federal oversight system and insufficient patient safety net protection.
    The earliest case of gene therapy gone astray was in 1999 and involved the University of Pennsylvania Institute for Human Gene Therapy, Dr. James M. Wilson and a teenager, Jesse Gelsinger. Jesse, aged 18 with a metabolic disorder of partial ornithine transcarbamylase deficiency, died of multiple organ failure after gene therapy of ornithine transcarbamylase attached to an attentuated adenovirus was infused into his liver. Again there was insufficient informed consent, IRB problems, and unreported adverse reactions. Arthur Caplan, the bioethicist at University of Pennsylvania deplored the privatisation of science, financial conflicts of interest, trade secrets.
    It appears that gene therapy is still in its infancy with insufficient patient protection/patient advocates in the face of economic opportunity for the companies’ personnel.

  10. Anonymous says:

    Sounds like this stuff would be an extremely wicked poison.

Comments are closed.