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="http://pipeline.corante.com/archives/clinical_trials/">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.