Icariin from horny goat weed is structurally unrelated to sildenafil (Viagra®)

In our last post, we discussed a recent USPTO ruling that rejected a claim of the Pfizer patent on the erectile dysfunction drug, sildenafil (Viagra®), to a novel oral treatment for the disorder. The patent appeals panel ruled that the existence of horny goat weed, a traditional Chinese medicine used orally for impotence, was grounds for rejection of the claim.

Frequent commenter daedalus4u pulled out US Patent #6,469,012 (filed Mar 4 1996, issued Oct 22 2002) wherein the relevant claim 24 reads:

A method of treating erectile dysfunction in a male human, comprising orally administering to a male human in need of such treatment an effective amount of a selective cGMP PDE inhibitor, or a pharmaceutically acceptable salt thereof, of a pharmaceutical composition containing either entity.

The claim should never have been allowed in the first place, said daedalus4u, and if upheld would have allowed Pfizer to block sales of horny goat weed. I'm also assuming that upholding claim 24 would have also allowed Pfizer to block sales of follow-on PDE5 inhibitors such as Lilly's Cialis®. But again, I'm not a patent attorney so I welcome further discussion on this point.

We also heard from my dear old colleague from Adelaide, Professor Ian Musgave. Ian is a pharmacologist who is probably better known in the blogosphere for his astronomy expertise (Astroblog, Southern Sky Watch, and contributing editor to Sky & Space).

But Ian is first and foremost a pharmacologist and raised this point:

Wait a minute, let me get this straight. The patent on Sildenafil, a compound that was rationally designed to specifically inhibit PDE (1) and developed prior to 1996 was invalidated because of a structurally unrelated component of a herbal extract that wasn't even shown to act on the cGMP/NO system before 2001? (2)

Pfizer is supposed to have time travel? The fact that a herbal treatment alleged to help sexual function would turn out, 5 years after the sildenafil patent had been granted (and the concept of PDE5 inhibitors as treatment for erectile dysfunction wide spread) to have effects on PDE5 could not possibly count as prior art. If so, the 1983 demonstration that the PDE inhibitor papaverine could produce erections would have invalidated the sildenafil patent.

The patent boards ruling is patently rubbish. How can sildenafil "logically flow" from a compound that wasn't even known to really work, let alone known to inhibit PDE?

(1)Nicholas K. Terrett, Andrew S. Bell, David Brown and Peter Ellis. Sildenafil (VIAGRATM), a potent and selective inhibitor of type 5 cGMP phosphodiesterase with utility for the treatment of male erectile dysfunction. Bioorganic & Medicinal Chemistry Letters
Volume 6, Issue 15, 6 August 1996, Pages 1819-1824
(2)Xin ZC. Kim EK, Tian ZJ, Lin GT, Guo YL. Icariin on relaxation of corpus cavernosm smooth muscle. Chin Sci Bulle 2001; 46: 485-9.

Prof Musgrave is indeed correct in the argument but daedalus4u points out the ruling did not invalidate the Pfizer patent, just the single claim to the pharmaceutical indication.

So that makes more sense to me and now I think I understand.

But in the discussion, Ian gave us another gem in response to a commenter who argued there was a structural resemblance between sildenafil and icariin, the active flavonol glycoside constituent of horny goat weed:

No, not at all. You can stick the benzofuran of the flavonol on top of the methylxanthine moiety of the sildenafil, but the charge distribution and 3D structure is all wrong, the flavonol has these huge sugar moieties which have the wrong charge and stick out in places that should inhibit binding, it lacks the bulky positively charged ring that fits into the PDE5 specific binding pocket. (I have the an image of the 3D structural overlay if anyone want to see them).

Looking at it, icariin does not scream either methylxanthine mimetic or "PDE inhibitor", in fact I would guess it would be worthless. However, flavonols and flavinoids have a habit of doing things we don't expect, such as resveratrol being a very decent cannaboid receptor anatgonist, despite looking nothing like cannabinoids of the synthetic antagonists.

In fact, Ian was kind enough to send us two molecular modeling images of the two compounds which I failed to get posted last night that illustrate the vast difference between sildenafil and icariin. This overlay image was generated with RASMOL, with icariin in the wireframe format and sildenafil in stick format:

Sildenafil Icariin RASMOL alignment_edited-1.gif

We reported yesterday that icariin has only 1/80th the potency of sildenafil in inhibiting PDE5 and its maximal effect on cGMP accumulation in an rabbit tissue model is 1/10th that of sildenafil. One doesn't need to be a computational chemist to see from Ian's overlay why this difference in potency and efficacy exists.

Anyway, this has been an interesting scientific exchange on an issue that was largely only covered in the pharmaceutical business press. This also illustrates for me the fun of science blogging - we all come to the table, albeit the electronic table, with our own expertise, have a discussion and learn from one another, even if some of us are halfway around the world.

More like this

Hey! Just a minute here. Are you seriously trying to apply science to statements from a guy who's just trying to make a quick buck off of something which sounds to the public as if it might work? Charge distribution, indeed!

By bigjohn756 (not verified) on 19 Feb 2010 #permalink

Thanks for the extended analysis of this issue Abel and elevating me to prominence!

Now I am not a patent lawyer either (even though I am an author on a patent, Method of controlling damage mediated by alpha, beta-unsaturated aldehydes, which has singularly failed to bring my co-authors and I fame fortune and research funding), but I still cannot see how, for the life of me, the fact that horny goatweed is used to treat erectile dysfunction has any bearing on claim 24.

Until 2001, 5 years after the patent was lodged there was no evidence that either horny goatweed or icariin had any effect on the NO/cGMP system, let alone the PDE5 enzyme.

As an example, look at the herb Yohimbe, it's used as a treatment for erectile dysfunction, and you could say the Pfizers' patent clause 24 was invalid because Yohimbe treats erectile dysfunction. But Yohimbine, the active ingredient of Yohimbe, acts in a completely different manner (it's an alpha-adrenoceptor blocker).

The point is that clause 24 is a clause about selective cGMP PDE inhibitors, not "anything that can treat erectile dysfunction in anyway imaginable". You can quite legitimately say clause 24 is too broad (but Pfizer hasn't stomped on any other selective cGMP inhibitors like tadalafil or vardenafil etc.). But that's the argument that should be used, not a spurious appeal to a herbal extract whose mechanism of action wasn't known even in outline until 5 years after the patent was filed.

There is a term that is sometimes used in patent lingo when someone puts stuff in their patent application that they don't need to, and which causes difficulties for people in the future trying to patent stuff. The term is called "crapping on the art".

In order to "crap" on the patent art, you don't need peer-reviewed stuff, you don't even need stuff that is accurate or even correct. Horny goat weed doesn't even need to work to treat erectile dysfunction, so long as it is believed to work, and is used for that indication, and is in use prior to the filing of the viagra patent application, and contains a selective phosphodiesterase inhibitor, then it counts and the claim must be rejected.

The mechanism of its function does not need to be known before the viagra patent filing, just that it is in use.

There are some real weaknesses and loopholes in patent law. One of them, that by demonstrating a single case you get to claim all cases is (I think) very damaging. Some of the early GM patents were like this, they showed one GM soybean, so they claimed all GM soybeans.

Meh, I sent in a long comment and the intertubes appear to have eaten it, after Science Blogs robot assuring me it had been received. I'll try again later.