Pharma’s Cutting Edge

The AOP of the July Nature Medicine contains the article “Neuropeptide Y acts directly in the periphery on fat tissue and mediates stress-induced obesity and metabolic syndrome” by Kuo et al from Georgetown U.  I’ve only read the abstract and stories so far (like this one from the Scientist and the press release from GU), but it’s enough to have me feeling elated at the discovery.  Congratulations to the Georgetown group.

Recall my post back in October 2006, when I used the failure of orally administered MK-0557, Merck’s NPY anatagonist, to discuss the state of pharma R&D knowledge transfer.  At the time, I implied that the development of NPY antagonists was all but dead.  Seems I may have spoken too soon.  Actually, I stick by my call regarding orally adminstered NPY antagonists, but if Kuo et al’s discovery of the role of NPY in glucocorticoid-mediated fat accumulation holds up to further scrutiny, I’ll be happy to revise that pronouncement as follows:  NPY anaagonists will have no role as systemically administered drugs aimed at influencing feeding behavior and thus obesity, but they might have a role as localy administered drugs aimed at controlling accumulation of local fat depots. 

The stories linked above describe some of the potential therapeutic uses of local NPY antagonist (and agonist) therapy.  Cosmetic uses are likely to be the first attempted, as it’s easy to approach the subcutaneous space and assess drug response.  Later uses could include non-surgical ablation of visceral fat (N.B. the omental fat dual-source blood supply may approached transcutaneously or intrarterially by experienced vascular interventionalists), which could very well treat metabolic syndrome. 

Intriguingly, activation of the peripheral NPY system may explain in full or part why glucocorticoids cause accumulation of fat in selected anatomical locations, including viscerally.  Although this explanation isn’t needed to enjoy the benefits of blocking the regeneration of cortisol from cortisone (via inhibition of 11-betahydroxysteroid dehydrogenase 1), as is being attempted with AMG221, Amgen’s 11 HSD1 inhibitor, it might help explain a failure of this or a similar drug to have the desired effects on visceral fat.  For instance, CT-guided skinny needle biopsies of visceral fat may be used to determine adipose tissue NPY and NPYR expression in the presence and absence of 11HSD1 inhibitor.  As far as I know, Amgen hasn’t released efficacy findings with this drug, and I’m not suggesting that it will fail to have desired clinical benefits.  Of course, the NPY mechanism might also be used to judge the effectiveness of such drugs before they ever get into humans too–a no-go for drugs that don’t block NPY upregulation in stressed mice fed ad libitum? 

Bottom line; this newly discovered peripheral NPY mechanism has great promise in metabolism drug discovery, as both a biomarker and as a drug target. 

As you’ll read in Therapeutics Daily and elsewhere, Novartis reported the delay of FDA’s review of Galvus, its answer to Merck’s recently approved Januvia, by three months.  Novartis said it submitted additional clinical data, triggering the FDA to add time to its review-cycle clock. 

What interests me about the story is the reasoning behind Novartis’ move.  Apparently, in  one or more monkey tox studies there was evidence of “skin lesions” that were of some concern.  Presumably, Novartis was trying to head off a dreaded approvable decision or maybe a nasty label warning with the additional clinical-data submission.  The nature of the skin lesions wasn’t elaborated by Novartis, although I read somewhere that blisters were involved. 

It’s hard to deduce too much from this limited information, but what I can tell you is that monkey studies aren’t run to look for skin toxicity specifically, so if a skin finding was made in monkeys, it was made during the course of a general toxicology or safety pharmacology study (more likely the former).  Also, Novartis wouldn’t have run a dermatotoxicity study (where the drug is applied to the skin) for an oral medicine like Galvus, so these skin lesions, whatever they are, were seen after oral administration.

Skin toxicity after oral drug administration is no rarity.  All sorts of skin reactions can occur with systemically administered drugs.  Usually such reactions are immune-mediated and most commonly are allergic reactions, including systemic contact dermatitis.  What is quite unusual about this case, though, is that drug-related skin reactions occurring after systemic administration, immune-mediated or otherwise, are not usually not picked up during preclinical testing at all.  They are usually first reported during clinical trials as human adverse reactions, sometimes not seen until after approval.

I couldn’t think of another example of a drug-related skin lesion following oral administration to monkeys, or any other non-human species, that was first seen during preclinical studies to compare with this case.  If you know of one, please enlighten us.

For their part, Merck reportedly denied the existence of any skin tox findings with Januvia.  Assuming that’s true, the Galvus monkey finding would appear to be specific to the drug and not an effect related to inhibition of DPP-IV per se.  GSK, whose DPP-IV Redona (denagliptin) was recently put on hold in Phase 3 for undisclosed toxicity issues, didn’t comment on the Novartis news.  Stay tuned for more from me on this story.

I’ll add this one into my vast collection of “I had no idea” new items. As someone who performed experiments, and worked closely with many other scientists who performed many experiments, on rodents fed commercial rodent chow, I have to say that I don’t recall a single discussion about the contents of the chow, outside of caloric and fat content. Phytoestrogens?! And I’m not talking about just any experiments, I’m talking about experiments in models of breast cancer and osteoporosis, conditions where the presence of pharmacological concentrations of dietary estrogens, albeit weak phytoestrogens, would certainly be of interest. It’s strange, now that I think about it, because we obsessed about the presence of weak estrogens in cell media–I distinctly remember how Katzenellenbogen’s discovery that Phenol Red was a weak estrogen immediately altered the cell-feeding practices of the entire breast-cancer research field. So why didn’t we obsess over the presence of weak estrogens in the diet of test animals? Strange that. I’ll be interested to learn more about the range of soy-derived phytoestrogens in rodent chow: Was it really enough to influence findings? What about other types of chow: dog and monkey in particular? What facts have been established that must now be called into question?

Another twist to the tragic story of the Tegenero CD28 agonist Phase 1 trial in which six healthy volunteers were strciken with severe adverse reactions that will leave them recovering for months. In this Times piece, Professor Angus Dagleish questions the decision to allow the trial, given prior experience with another monoclonal antibody that, like the Tegenero drug, increased T cells via co-stimulatory pathway modulation and also led to serious side effects. He also questions the decision to proceed with administration to six volunteers simultaneously. Expect to see more of the Monday-morning quaterbacking in the weeks and months ahead, as formal investigations and, probably, legal battles play out. Generally speaking, administration of an investigational drug to six volunteers simultaneously is not unusual, since the initial dosing is always begun well below the anticipated toxicity level based on preclinical data. What is quite unusual is the failure of preclinical studies to predict severe toxicity at the starting dose. Why this failure now? I can’t speculate wisely without additional information. Assuming the predictive failure was the result of a failure to identify a pharmacological effect (i.e. related to the drug’s mechanism of action), as opposed to a pharmaceutical effect (i.e. related to the pharmaceutical properties of the drug per se), it represents the inability to quantify either the degree or implications of cell-mediated immunity over-drive via co-stimulatory pathway agonism. In that case, all molecules that work through this mechansim would be expected to suffer from the same imprecision of preclinical testing, necessitating extreme caution in their application, as advised by Professor Dagleish.

From a broader lessons standpoint, this tragedy serves as a jolting reminder of the need for much better predictive science prior to human testing, a need the Critical Path Institute and FDA are addressing via their proposed predictive safety consortium initiative that facilitates sharing of information among major pharmaceutical labs.