- The compound, or composition of matter, patent for Revlimid (US 5,635,517) expires in October 2019.
- Celgene has a polymorph patent for Revlimid (US #7,465,800) that expires in 2027.
- Given the Lipitor (atorvastatin) precedent, the Revlimid polymorph patent, even if upheld, will likely not prevent introduction of generic Revlimid in 2020.
- Two method of use patents (7,189,740 and 7,968,569), which expire in 2023, may become the battle ground for extending exclusivity for Revlimid beyond 2019.
Let me start with a disclaimer: I am not a patent lawyer, nor a scholar in polymorphs and formulations, and I also have no position in Celgene (NASDAQ:CELG). I am writing this article primarily to generate more discussion from a wider audience of experts on the likelihood of Celgene maintaining exclusivity for Revlimid beyond their compound expiry in Oct 2019. This article builds off of an excellent video article that appeared recently in Seeking Alpha on this topic and readers are referred to that article for additional background.
Revlimid (lenalidomide) accounted for 67% of Celgene's 2013 revenue ($4.3 B) and is expected to generate revenue of $7 B by 2017. Since CELG is launching additional drugs in the next few years, Revlimid as a percentage of sales should decrease somewhat, but will still account for 50% of revenue in 2019. Thus, introduction of lower cost generics in the US would have a huge impact on Celgene's revenue starting in 2020.
Given the importance of Revlimid to the future of the company, CELG has built a fortress of patents around this drug - 11 are orange-book listed. It will therefore be hard for generic companies to navigate around the numerous patents in order to market a generic version. However, given the huge sums of money at stake, they are already gearing up for the fight. The first shot across the bow has come from a joint effort between Natco and Actavis, who have teamed up to challenge several of the CELG patents. Quite a bit of interest has been generated by this challenge, which will be heard in a Markman hearing in mid-May (for more details, see here).
Compound, or composition of matter, patent US 5,635,517
Expert analyses from Barclays and Cowen & Co. generally seem to agree that the main compound patent, US 5,635,517, expiring in October 2019, is very solid and unlikely to be rendered invalid. In Europe, this patent expires in 2022 and Cowen & Co expects generic entries in Europe after that date.
Polymorph Patent, US 7,465,800
The patent of most interest is the so-called "polymorph" patent, US 7,465,800, which expires in 2027. Analysts appear to be largely in agreement that this patent will also be upheld. Barclay's comments "we believe Celgene has a strong case with its polymorph patent (and other patents) based on prior case judgments."
The '800 patent describes 8 polymorphs, denoted forms A through H, which include three unsolvated forms, the hemihydrate, a dihydrate, an acetone hemisolvate, an acetonitrile solvate, and a dehydrated form of the dihydrate. However, only the hemi-hydrate form (containing one-half mole of water) is claimed. This form is therefore likely different from the material isolated in the '517 patent, which was isolated from an anhydrous solvent system and should be free of water. In the '800 patent, Celgene does not describe any specific advantages (such as improved bioavailability, stability, ease of formulation) of using the hemi-hydrate form vs. the form generated in the '517 patent. This is usually required for polymorph patents these days but apparently was not required for this patent.
Assuming the polymorph patent is upheld, both the Barclays and Cowen & Co. analyses point out that generics could "work around" the polymorph patent by (1) simply practicing the '517 patent once it expires in 2019 or (2) developing a new polymorph, salt, or amorphous form that is not covered in the CELG patent and would therefore be non-infringing. In essence, upholding the Revlimid polymorph patent would make it harder to introduce a generic form, but not impossible.
Practicing the '517 patent is the first approach a generic company could explore. In the '517 patent, the drug is crystallized from ethyl acetate/dioxane, in the absence of water, so it is likely to be an anhydrous polymorph. No solid-state characterization of this polymorph is provided in the '517 patent other than the melting point. A generic company could practice the '517 patent to generate a form that has the same melting point as that described in the '517 and does not contain water so it clearly does not infringe the '800 patent, and would have freedom to practice. In fact, a generic company could use any of the non-claimed forms described in the Celgene '800 patent. The challenge for a generic company will be to develop a shelf-stable formulation that is bioequivalent to the CELG formulation. This could prove to be demanding since CELG reports their formulation is >90% bioavailable. However, given the long runway for generic companies to work on development of a viable formulation, the odds are high that they will be able to do so.
The second approach would be to use amorphous drug, as noted in the Barclays report. The amorphous form of the drug is not patented. The generic company would simply have to show that their formulation contained none of the crystalline hemi-hydrate and would therefore be non-infringing to the '800 patent. A number of new developments in formulation technology have now made amorphous formulations viable, including spray-drying and hot-melt extrusion. These types of formulations work best on low dose drugs, which means they would be feasible for any of the doses of lenalidomide (doses range from 2.5 mg to 25 mg).
Precedent for use of amorphous formulations to non-infringe a polymorph patent comes from generic introduction of Lipitor (atorvastatin) once Pfizer's compound patent expired in Nov 2011. Pfizer's patent (5,969,156) on atorvastatin polymorphs and salt forms is still active (expires in Jan 2017) yet generic atorvastatin was approved soon after the compound patent expired in the US. In 2009 and 2010, Pfizer sued Mylan, Kremer-Urbans, Aurobindo, Actavis, and Apotex, all of which had filed ANDAs to market generic atorvastatin. However, the suits were futile and today more than a dozen companies market generic atorvastatin in the US. Meanwhile, Pfizer's sales of Lipitor dropped to $2.3B in 2013 after peaking at $11.4B in 2009.
The third approach to work around the '800 patent is to develop a non-infringing salt form. Lenalidomide is only weakly basic (pKa = 2.3), but crystalline acidic salts potentially can be formed. In fact, a US patent application (20120190711 A1) has already been published that claims the benzenesufonate and toluenesulfonate salts, which have the advantage of being more water soluble than the free base. Other salts are discussed in the patent but are not claimed. If granted, the applicant would have IP on their generic version of lenalidomide. Even if not granted, formulations of these salts would be non-infringing to the Celgene patent.
Gleevec as a harbinger for Revlimid
In addition to the atorvastatin precedent, Gleevec (imatinib mesylate) should serve as a good example of what we might expect for Revlimid. The Gleevec compound patent in the US expires in Jan 2015 but Novartis has a polymorph/salt form patent on imatinib mesylate that expires in May 2019. The generic company Hetero has already been granted a patent on alternate polymorphs of imatinib (including hydrates), US 7300938, and may be poised for generic introduction early next year. Sun Pharma filed an ANDA way back in 2007 and in 2013 sued Novartis to invalidate their polymorph patent. In an article published by Bloomberg in late 2011, Novartis management seemed resigned to generic introduction in the US in 2015. Novartis CEO Joseph Jimenez discussed their strategy to transition use from Gleevec to their improved drug, Tasigna. In 2013, Tasigna had sales of $1.3B and was growing 30% annually while Gleevec had sales of $4.7B with no annual growth vs. 2012. In any event, it seems surprising that we have not seen more activity from generic companies other than Sun Pharma at this late date.
Revlimid Method of Use Patents
The '517 patent, which expires in 2019, covers methods for using lenalidomide to reduce undesirable levels of TNF alpha but not for treating specific diseases. CELG owns two method of use patents in the US for Revlimid for treating specific cancers:
· 7,189,740 for use as a single agent or in combination with other cancer drugs for the treatment of myelodysplastic syndrome, expiring in Apr 2023
· 7,968,569 for use as a single agent or in combination with dexamethasone for multiple myeloma, expiring Oct 2023.
These patents have not generated much attention but they may become critical since it seems likely that generics will try to enter the US market with non-infringing formulations in 2020. The current challenge from Natco/Actavis on these two patents appears to be focused on wording technicalities, which never seems like a strong approach. Barclays thinks Celgene might have the edge with one while Natco/Actavis might have the upper hand in the other.
As Rena Conti has pointed out in a Dec 2013 article, previous case law has not generally supported method of use patents as a barrier to entry. In April 2012, the U.S. Supreme Court and the U.S. Court of Appeals for the Federal Circuit ruled in 2 cases that the innovator companies could not prevent regulatory approval of generic products based on their use patents. The court ruling indicates that one patented use will not prevent marketing a generic drug for other unpatented uses. However, the court commented that once generic entrants are available in the market, branded manufacturers can monitor the generic's marketing programs to make sure they do not infringe the method of use patent. In other words, generic companies can obtain regulatory approval for the drug but cannot market it for the uses that are still under patent. However, once the generic drug is approved, doctors would be free to prescribe a generic drug regardless of the patent situation.
Revlimid is approved for 3 indications:
· Myelodysplastic syndrome
· Multiple myeloma
· Mantle Cell lymphoma.
If my understanding of the method of use patents is correct, only the first two are covered, so potentially a generic company could file for the unpatented indication in order to obtain ANDA approval. If all three uses are in fact patented, then these patents could be a barrier to introduction of generics until 2023 in the US.
Alimta - method of use case study
The Alimta case provides further perspective on method of use patents although the circumstances are different from the Revlimid case. Alimta's composition of matter patent expires in early 2017 but Lilly has a use patent, which extends to 2022. In the method of use, a combination of vitamin B12 and folic acid is given to patients prior to Alimta administration to reduce treatment-related hematological and gastrointestinal toxicity, which can be serious and sometimes fatal. As reported in Fierce Pharma, Teva challenged the use-patent based on obviousness but in late Mar 2014 the judge ruled in favor of Lilly since the scientific discovery of the correct use of B12 and folic acid to avoid toxicity while maintaining activity against cancer was not obvious and therefore patentable. Since Alimta must be used by this protocol, it is likely that generic entrants will be delayed.
Analysts agree the composition of matter patent '517 is likely secure through its expiry date of Oct 2019.
While analysts have focused on the polymorph patent '800, which expires in 2027, precedent from atorvastatin suggests this patent will not be a barrier to entry of non-infringing formulations starting in Oct 2019. In addition, Gleevec, with a compound patent expiry in Jan 2015 and polymorph patent in 2019, will provide further perspective on what to expect for Revlimid.
The two method-of-use patents, which expire in 2023, will likely be the main focus of litigation as the end of this decade nears. It is probable that generic companies will file ANDAs for a non-patented indication, with non-infringing formulations, at least 30 months prior to the expiration of the '517 patent. ANDA filings 30 months prior to patent expiry are likely since it can be expected that CELG will sue the generic companies. Under the Hatch-Waxman Act, once Celgene sues for infringement within 45 days, the FDA will not grant the generic company final approval for 30 months, unless the court rules the patent is not infringed, invalid or unenforceable prior to that date. It seems likely that the courts will play an integral role in determining when generic Revlimid will enter the US market.