About this Author
DBL%20Hendrix%20small.png College chemistry, 1983

Derek Lowe The 2002 Model

Dbl%20new%20portrait%20B%26W.png After 10 years of blogging. . .

Derek Lowe, an Arkansan by birth, got his BA from Hendrix College and his PhD in organic chemistry from Duke before spending time in Germany on a Humboldt Fellowship on his post-doc. He's worked for several major pharmaceutical companies since 1989 on drug discovery projects against schizophrenia, Alzheimer's, diabetes, osteoporosis and other diseases. To contact Derek email him directly: Twitter: Dereklowe

Chemistry and Drug Data: Drugbank
Chempedia Lab
Synthetic Pages
Organic Chemistry Portal
Not Voodoo

Chemistry and Pharma Blogs:
Org Prep Daily
The Haystack
A New Merck, Reviewed
Liberal Arts Chemistry
Electron Pusher
All Things Metathesis
C&E News Blogs
Chemiotics II
Chemical Space
Noel O'Blog
In Vivo Blog
Terra Sigilatta
BBSRC/Douglas Kell
Realizations in Biostatistics
ChemSpider Blog
Organic Chem - Education & Industry
Pharma Strategy Blog
No Name No Slogan
Practical Fragments
The Curious Wavefunction
Natural Product Man
Fragment Literature
Chemistry World Blog
Synthetic Nature
Chemistry Blog
Synthesizing Ideas
Eye on FDA
Chemical Forums
Symyx Blog
Sceptical Chymist
Lamentations on Chemistry
Computational Organic Chemistry
Mining Drugs
Henry Rzepa

Science Blogs and News:
Bad Science
The Loom
Uncertain Principles
Fierce Biotech
Blogs for Industry
Omics! Omics!
Young Female Scientist
Notional Slurry
Nobel Intent
SciTech Daily
Science Blog
Gene Expression (I)
Gene Expression (II)
Adventures in Ethics and Science
Transterrestrial Musings
Slashdot Science
Cosmic Variance
Biology News Net

Medical Blogs
DB's Medical Rants
Science-Based Medicine
Respectful Insolence
Diabetes Mine

Economics and Business
Marginal Revolution
The Volokh Conspiracy
Knowledge Problem

Politics / Current Events
Virginia Postrel
Belmont Club
Mickey Kaus

Belles Lettres
Uncouth Reflections
Arts and Letters Daily
In the Pipeline: Don't miss Derek Lowe's excellent commentary on drug discovery and the pharma industry in general at In the Pipeline

In the Pipeline

« Arzoxifene: Not the Road to Big Profits? | Main | Ariad's Patent Rises From the Grave to Smite the Living »

August 25, 2009

Polymorphs and Salts: India Raises an Eyebrow

Email This Entry

Posted by Derek

As some of you may know, there's a big patent dispute between Novartis and the government of India. The issue is Gleevec (imatinib, sold as Glivec in most of the rest of the world - Novartis must have figured that it would have been pronounced "Gly-veck" over here). The product is sold as a mesylate salt, and in fact, as a particular polymorph of that mesylate salt, and there's the problem.

For those outside the business, most drugs have either acidic or basic groups on them, and you can make a salt of them by combining them with a corresponding base or acid. Basic drugs - amines, mostly - are often sold as hydrochloride, mesylate, citrate, etc. salts, and acidic drugs are often sodium, potassium, calcium, etc. salts. These changes are usually done to make a compound absorb better when it's dosed and/or to make it easier to handle or more stable during manufacturing and storage.

Polymorphs, meanwhile, are different crystalline forms of the same compound. That's something that you don't encounter much outside a chemistry lab. The closest everyday analog is to think of table salt vs. kosher salt vs. sea salt, but those are still the same crystal-packing form when you get right down to it. A real polymorph is quite a different beast; it's as if you could dissolve up regular salt, cool it down in some tricky way, and have it crystallize out as needles or prisms instead of tiny cubes. And those needles or prisms might then, as it happens, refuse to dissolve if you added them to your soup. That's a polymorph, and it's a pretty common occurrence with drug substances. A key step in a real manufacturing process is making sure that you have the best one, and that you can always be sure that it's the one being produced. The wrong one will do things like refuse to dissolve into the bloodstream, which can be most unfortunate.

So Gleevec is a particular polymorph of a particular salt, and Novartis has patents on just that form in many countries. But not India, or not yet. As this post from a lawyer there details, the dispute is (to a large extent) about whether this form of the drug should be compared to another polymorph, to another salt, or to the original free base compound when time comes to judge its novelty and patentability. Another question is whether Novartis's previous patent filings disclose or anticipate the particular salt and polymorph form of the final compound. These arguments are complicated by the fact that India didn't even allow patents on pharmaceutical substances until a few years ago. For more on recent drug company patent disputes there, see this from the WSJ.

So I'd like to throw a question out to the readership: how many examples can people think of where a particular salt or polymorph was a key to getting good efficacy or properties for a drug? I realize that a lot of these stories never see the light of day - I've seen polymorph problems give people fits during development, as have many readers, I'm sure, but most of these things never get published. So I'm not asking for anything from the inside, just the publicly known examples.

Update: if you want a good indicator of how serious the IP issues are around these things, check out this conference. . .

Comments (35) + TrackBacks (0) | Category: Drug Development | Patents and IP


1. Anonymous on August 25, 2009 8:53 AM writes...

Great post Derek. Much like you I can think of dozens of unpublished examples from inside. The infamous example of a polymorph problem seriously screwing things up is, of course, Abbott's Ritonavir.

(for those of you that don't know a more stable (ie less soluble) polymorph of Ritonavir was discovered whilst the drug was being manufactured. Its appearence meant the polymorph under development could not be made reproducibly, and the product had to be withdrawn from market while the drug was re-formulated costing Abbott some serious $)

Permalink to Comment

2. petros on August 25, 2009 8:59 AM writes...

Crystalline forms have proved very valuable to marketed drugs.

Glaxo got several years extra patent life through the Form 2/Form 1 saga of rantidine. Since the Form 2 patent expired much later than the Form 1 it was able to block most generics becuase of the problems of obtaining pure Form1 material

Permalink to Comment

3. barry on August 25, 2009 9:00 AM writes...

The example that leaps to mind is Abbott's ritonavir. They had an approval from the FDA in hand when a new--stabler, less soluble--polymorph showed up. Dissolution rate (and bioavailability) was much lower, and it was clogging up their equipment in the manufacturing facility. A large effort by many people eventually figured out how to get one or the other polymorph at will, but the expense was large.
Glaxo went to court to defend a patent on a second polymorph of cimetidine after the patent on the first polymorph had expired.

Permalink to Comment

4. Jose on August 25, 2009 9:28 AM writes...

Advair, for sure. Xinafoate was certainly not the first salt GSK monkeyed around with.

Permalink to Comment

5. Vader on August 25, 2009 9:41 AM writes...

Probably only of historical interest, but my recollection is that finding a suitable salt of penicillin was a big part of making it a practical drug.

Permalink to Comment

6. addy on August 25, 2009 9:43 AM writes...

another common eg in textbooks is of chloramphenicol palmitate, with 3 polymorphs-A,B,C.

Permalink to Comment

7. CMCguy on August 25, 2009 9:50 AM writes...

The Abbott HIV drug and the Glaxo Acid Reflux inhibitor mentioned are dramatic examples of manufacturing and IP consequences of polymorphs and I think prior to those "events" much less attention was generally paid to such possibilities. I worked on dozen's of oral products spanning period where one used to check quickly if any different crystal forms might occur as a secondary matter (although found a few places already in serious mode as SOP) then became a critical ingrained part of the development process to define (at least at most places). There were only a couple cases I saw where it seemed to matter in terms of dissolution rate so normally would target form that was most reproducible assuming the physical characteristics amenable to manufacturing. Of all the headaches inherent in Drug development this can be a big challenge as even with systematic studies FormDev seems more art than science IMO.

Permalink to Comment

8. Doubja on August 25, 2009 9:52 AM writes...

Pfizer recently reported that they came across a polymorph for one of their high-profile HCV inhibitors which led to its deprioritization and another JMC paper:
J. Med. Chem., 2009, 52 (5), pp 1255–1258

Permalink to Comment

9. on August 25, 2009 9:53 AM writes...

It's not a drug example, but polypropylene is known to have 3 crystalline forms. The alpha is more common, but the beta can be induced with proper nucleating agents. The thermal and mechanical properties are not the same, with the beta having a lower melting point and greater impact resistance.

Permalink to Comment

10. Anonymous on August 25, 2009 9:59 AM writes...

CMCGuy - 'There were only a couple cases I saw where it seemed to matter in terms of dissolution rate...'

Wow, you got very lucky my friend!

Permalink to Comment

11. SSS on August 25, 2009 10:56 AM writes...

From Syntex - Naprosyn (free base) and Anaprox (Na salt) - 2 separate drugs, separate NDAs. Aleve (OTC) is the sodium salt. Faster dissolving, better %F.

Permalink to Comment

12. milkshake on August 25, 2009 11:11 AM writes...

I heard of a case of a drug candidate where the re-crystallization into a new (more stable) polymorph was induced unpredicatably by pressing the tablets, which then led to some tablet batches to fall apart, and the formulation work and all animal and pilot human studies had to be re-done for the new polymorph formulation.

There was a paper in Org Process R&D recently and the authors wanted to be sure they won't get a nasty surprise like this one down the road so they left a batch of the material stirring for like two months - and sure enough eventually they obtained a more stable salt polymorph which then they always used for seeding when re-crystallizing their compound.

Whats really obnoxious is that sometimes the polymorph formation is facilitated or inhibited by trace quantities of impurities (which can vary), and every time the preparative scale and equipment is changed there is chance for unpleasant surprise.

Permalink to Comment

13. Salt Selector on August 25, 2009 11:26 AM writes...

The Ca salt of atorvastatin (Lipitor) is significantly better than the other available forms, because it prevents cyclization of the molecule and provides long term stability + enhanced solubility/oral absorption in vivo. It is apparently really difficult/next to impossible to formulate other salt forms/free form/amorphous form not covered by their patent and still maintain stability + exposure levels. Its not clear whether generic companies will have that level of expertiese/luck. In this case, the Ca phase patent may extend the patent life past the composition of matter patent, which could mean continued profits for Pfizer.

Permalink to Comment

14. Development Guy on August 25, 2009 12:06 PM writes...

Slightly related: Kurt Vonnegut's book "Cat's Cradle" explored the idea of the discovery of a new polymorph of water, one that's a solid at room temperature. Very interesting, very bad things happen...

Permalink to Comment

15. milkshake on August 25, 2009 12:44 PM writes...

the unlikelihood of thermodynamically more stable version of ice (not arising spontaneously over 4+ billion years) is what left me cold to the Vonnegut's premise. But tin pest is known and rather dramatic demosntration of the allotrope problem

Permalink to Comment

16. Harry on August 25, 2009 1:19 PM writes...

Not really a pharma tale of polymorph horror, but the place I used to work produced benzyl phenols.

The process produced a mixture of ortho- and para- benzyl phenols, as well as some small amounts of di and tri benzylated material.

We separated the ortho (desired) isomer from the rest by vacuum distillation, with the orth- being the lower-boiling isomer.

We had produced this product for many years, and the ortho always came out as a colorless liquid, with a melting point around 20 C.

We were aware that there was another crystalline form, that had a melt point of around 51-52 C, but we never encountered it until one day, for no apparent reason the product in the receiver crystallized. The small amount of tracing we had on the receiver was not enough to melt it, and we ended up having to use a couple of steam hoses to empty the receiver.

From that day on, we were never able to obtain the low-melting material. Fortunately, this didn't matter to our customer, but it made it really tough operationally (we ended up putting a clamp-on steam jacket on the receiver).

That's my tale of woe. It's amazing how things will jump up to bite you in this business.

Permalink to Comment

17. CMCguy on August 25, 2009 1:41 PM writes...

#10 Anon- I don't disagree although attribute my "luck" to two factors, small sample size and working with others who guided me well. I heard many horror stories but in my limited experience never encountered a true head knocker. I also was fortunate to work with a few Form Dev experts that had a innate sense of what to do/not do (including if greater probability of polymorphs). Of course those same experts were the ones that told me about big problems they had on projects which is likely how they gained the necessary senses.

Permalink to Comment

18. Curious Wavefunction on August 25, 2009 2:59 PM writes...

Tin pest. Think Napoleon's army's buttons.

Permalink to Comment

19. Artemisinin on August 25, 2009 4:12 PM writes...

"The wrong one will do things like refuse to dissolve into the bloodstream, which can be most unfortunate."

I am wondering why is bloodstream.

Permalink to Comment

20. ex-Pfizerite on August 25, 2009 5:47 PM writes...

the salts of Norvasc have a major impact on tablet formation and process flowabiliry

Permalink to Comment

21. weirdo on August 25, 2009 6:57 PM writes...


I assume he meant "refuse to dissolve in the stomach" (or "colon").

Permalink to Comment

22. researchfella on August 25, 2009 11:30 PM writes...

I've always thought that it was somewhat ironic, that after years of high-tech complex research including target validation, cell biology, pharmacology, medicinal chemistry, computational modeling, etc., etc., a simplistic thing like a new crystal form with a slightly higher melting point can de-rail (or almost de-rail) a program. I've experienced first-hand how frustrating that can be.

Permalink to Comment

23. milkshake on August 25, 2009 11:53 PM writes...

The almost-derailing is typically a management-created problem rather than irony. By the time you got a clinical candidate there is a already lot of pressure to "have a clinical candidate by the end of this year" that formulation work is often not given enough time and priority, and then of course you can have unpleasant surprises later on.

Fortunately there are now experts that offer to do polymorph screening for you. (And it is politically a lot easier to pay up 600k for a contract rather than hiring two extra formulation bench chemists to do the necessary work in-house).

Permalink to Comment

24. bcpmoon on August 26, 2009 1:26 AM writes...

Ahh- Polymorphs. Rather the rule than the exception. You should always have your fingers crossed when scaling up. I currently have a case where it is important to crystallize at a certain temp but please cool to that temp and do not heat. Of course in hindsight obvious but who thinks of everything when piloting and holding the solution overnight for technical reasons at 0°C? And most of the products have to be milled at a late stage and -presto- a wonderful world of opportunities for polymorph formation.
More an art than anything else, I agree.

Permalink to Comment

25. CMCguy on August 26, 2009 10:00 AM writes...

#23 milkshake while management mandated deadlines do create intense pressure as suggested, and not only to the formulation selection, this is a case where the chemists and R&D systems often share the responsibility for decisions/inadequate studies. Although as the post illustrates there is awareness of polymorphs as a potential issue med chemists and even those in process rarely pay attention early enough to salt forms and physical properties. The screening studies you mention are great and more routine these days although usually happen after a candidate identified rather than as a critical selection criteria. Some of this is the fact due to limited amounts of compounds making doing the work difficult but it also appears to be a mindset: formulations is someone else's problem. This mindset can be reinforced by organizational structures where indeed Form Dev can be isolated and not actively engaged soon enough, particularly with discovery efforts. Having formulations input during research can help keep project on the right track even though they can suffer similar surprises to other development groups.

Permalink to Comment

26. Don B. on August 26, 2009 10:13 AM writes...

A famous case is Dilantin(TM) (phenytoin sodium).

Parke Davis sold the product for grand mal epilepsy. The VA in their infinite wisdom purchased generic phenytoin sodium &nd had patients all over the country having grand mal seizures./

Permalink to Comment

27. Hap on August 26, 2009 2:20 PM writes...

#26: Isn't the point of a generic to be equivalent to the previously patented compound? If the drug wasn't, then wasn't either the generic drug company, the FDA (for not checking the drug's equivalence to the previously prepared drug or its consistency), or the patent office/original manufacturer (for not disclosing method) at fault? If generics weren't expected to work like the originals, then they would no longer be substitutable for the originals and they likely wouldn't be usable - so making the alternate assumption seems to make sense.

Permalink to Comment

28. Salt Selector on August 26, 2009 7:59 PM writes...

One other thing to consider with polymorphs is the path through development prior to market, rather than the final product. Often getting the exposures required for a manufactured product isn't nearly as tough as getting the high exposures required for Phase I Single Ascending Dose and Multiple Ascending Dose studies, or the crazy high exposure multiples required in preclinical SA studies. Newly discovered less soluble polymorphs can throw a wrench into these studies.

Permalink to Comment

29. Salt Selector on August 26, 2009 7:59 PM writes...

One other thing to consider with polymorphs is the path through development prior to market, rather than the final product. Often getting the exposures required for a manufactured product isn't nearly as tough as getting the high exposures required for Phase I Single Ascending Dose and Multiple Ascending Dose studies, or the crazy high exposure multiples required in preclinical SA studies. Newly discovered less soluble polymorphs can throw a wrench into these studies.

Permalink to Comment

30. milkshake on August 27, 2009 1:41 AM writes...

Generics are supposed to be equivalent but sometimes the generic manufacturer does not get the formulation quite right or the standards slip later on. I would be quite careful with generic sustained-release drugs - they may not work too well. There were recently problem with Indian companies fabricating their formulation test results for FDA submission. Also Teva in Israel has had some problems. Formulation is alchemy.

Permalink to Comment

31. Anonymous on August 27, 2009 3:14 AM writes...

Couldn't agree more with #25. It's not true of all discovery chemists of course, but many hold the mindset that 'development will fix that, it's not our problem'. In many respects, it's not the medicinal chemists problem (they have plenty of problems of their own, I am well aware of this!). However, a better understanding of polymorphism and it's potentially nightmarish consequences can only be a good thing for all chemists.

That in mind, thanks for the post Derek!

Permalink to Comment

32. processchemist on August 27, 2009 9:47 AM writes...


It's obvious, med chemists have problem on their own... It's mainly a project management issue. In recent times I've seen PM that put polymorph screening/studies in the Phase III process tecnology package, and I've seen troubles arising from the formulation side between Phase I and II...

Permalink to Comment

33. CMCguy on August 27, 2009 10:29 AM writes...

#32 processchemist I have encountered many PMs who don't comprehend drug development, particularly the CMC areas since if pharma experience normally backgrounds come from strictly clinical side. Typically PMs only parrot what they learn and have seen on other projects so again to me it comes back to the bench chemists and their direct managers providing correct info and indeed fighting to get studies/work done at appropriate time to do good.

#31 Anon is correct that med chemists have own problems and I am not advocating addition of a complete polymorph/formulation task to list. What I do suggest is that chemists, of all flavors, interact with Form Dev early on so perhaps can avoid missteps. I would expand that to many other group interactions as well because the more scientists work in isolation (silos- which can be part of academic training and industry organization) the less efficient their work is in the overall scheme of drug development, which is inherently fairly bumpy.

Permalink to Comment

34. Sili on August 29, 2009 12:04 PM writes...

I wish I could get to work with this. I think it's one of the most fascination aspects of crystallography - unfortunately so does everyone else.

Did encounter some interesting polymorphism in the coördination compounds I worked on for my (never finished) ph.d. Unfortunately I never got the physical charecterisation that'd allow me to use the structures for anything more that pretty pictures.

Permalink to Comment

35. Puhn on August 30, 2009 12:59 AM writes...

#25: I came across a web site of a small company claiming they do quantum-based multi-property in-silico lead optimisation. Solubility included. Not much more detail, but if what they say about their Malaria drug candidate - IC50 176 nM with TI>6900 (!) probably they can be helpful to the Form Dev guys too.

Permalink to Comment


Remember Me?


Email this entry to:

Your email address:

Message (optional):

The Last Post
The GSK Layoffs Continue, By Proxy
The Move is Nigh
Another Alzheimer's IPO
Cutbacks at C&E News
Sanofi Pays to Get Back Into Oncology
An Irresponsible Statement About Curing Cancer
Oliver Sacks on Turning Back to Chemistry