In-depth analysis of Tarceva patent litigation

Continuing from the last post that discussed background of Tarceva patent and the single judge order rejecting Roche's application for interim injunction to restrain Cipla from manufacturing and selling generic product Erlocip, here we discuss appeal filed by Roche before the divisional bench at the Delhi High Court against the rejection order. 

C. Appeal before the Divisional bench
After briefly taking note of the key arguments pleaded by Cipla and Roche before the single judge and further briefly reviewing the observations made by the single judge, the divisional bench picked up the issue of or rather convinced to zero in over Cipla’s “polymorphs A+B” argument. This is the same argument that Cipla’s counsel tactfully pleaded before the single judge but this time the impact made was considerably far-reaching to convince divisional bench to refer compound as claimed by the IN‘774 patent as “polymorphs A+B” rather Erlotinib hydrochloride per se.

The divisional bench while taking note of Cipla’s “polymorphs A+B” argument, first and foremost, weighed whether Roche even made a prima facie case for injunctive relief. The divisional bench unequivocally agreed with Cipla’s counsel that the IN‘774 patent is a mixture of polymorph A and B forms of Erlotinib hydrochloride and Tarceva product sold by Roche is polymorph B form of Erlotinib hydrochloride and further opined that Roche not even made a prima facie case for injunctive relief as Roche do not hold patent for polymorph B form of Erlotinib hydrochloride in India.

The divisional bench insistently opined that the grant of injunction ought to be refused solely on the ground that Roche suppressed the fact that Tarceva product is polymorph B form of Erlotinib hydrochloride and the IN‘774 patent is “polymorphs A+B” of Erlotinib hydrochloride both before the Controller of Patents and the court. Keeping “polymorphs A+B” reasoning in mind, the divisional bench further noted

• That Roche tried to mislead both the divisional bench as well as the Controller of Patents that polymorph B was subsumed in “polymorphs A+B”. The divisional bench even reasoned that if Tarceva product corresponded to “polymorphs A+B” then there was no need for Roche to file a separate application for polymorph B. In short, the divisional bench entirely disagreed with Roche that polymorph B is subsumed in “polymorphs A+B”.

• That Roche failed to brought to the notice of the Controller of Patents that the IN‘774 patent is for “polymorphs A+B” of Erlotinib hydrochloride while prosecuting the application, which the divisional bench reasoned was not consistent with the requirement of a full disclosure. The divisional bench even reasoned that if the Controller of Patents was cognizant of “polymorphs A+B” fact then the Controller would have taken it in account under Sec. 3 (d) for efficacy requirement. In short, the divisional bench noted that failure to brought “polymorphs A+B” fact to the notice of the Controller raised a more than credible challenge on the validity of the IN‘774 patent.

• That Cipla raised serious doubt whether Roche even hold a patent for Tarceva product and made a prima facie case for injunctive relief under proviso of Sec. 11A (7) of the Act which states “provided that the applicant shall not be entitled to institute any proceedings for infringement until the patent has been granted.” The divisional bench further noted that Roche is not commercially exploiting the IN‘774 patent since Tarceva product corresponds to polymorph B of Erlotinib hydrochloride and not “polymorphs A+B”. In short, the divisional bench noted that Cipla demonstrated prima facie that Roche is not entitled to enforce the IN‘774 patent and cited Franz Xaver Huemer v. New Yash Engineers (AIR 2000 Del 23) where the court held that the patent which is not commercially utilized cannot be enforced.

• That the Controller rejected the IN‘507 application for polymorph B of Erlotinib hydrochloride but never took into account facts pertaining to polymorph B while deciding the pre-grant opposition concerning the IN‘774 patent. In short, the divisional bench noted the Controller failed to take into account facts about polymorph B of Erlotinib hydrochloride while deciding the patentability of the compound “polymorphs A+B” which add credible challenge on the validity of the IN‘774 patent.

• That the single judge proceeded on the fact that Roche holds a valid patent for Tarceva product but in view of the fact that Tarceva corresponds to polymorph B form of Erlotinib hydrochloride, the divisional bench noted that Roche failed to make prima case and injunction ought to be refused.

The divisional bench even laid down “must disclose” conditions for plaintiffs in order to make prima facie case for interim injunction categorically for infringement of patent covering pharmaceutical drug. First, plaintiffs have to disclose the XRD data of the product and make full disclosure of the complete specification of infringed patent. Second, plaintiffs have to make unequivocal disclosure that the patent covers the pharmaceutical drug. Third, plaintiffs have to disclose any other pending applications and issued patents pertinent to derivatives or forms of the pharmaceutical drug. The divisional bench further opined that absence of any of “must disclose” details would be a case of suppression of material facts.

The divisional bench while weighing whether Cipla raised credible challenge to the validity of the IN‘774 patent noted:

• That Cipla has been able to show that the Controller’s order to grant the IN‘774 patent was arguably deficient of the fact that the closest prior art EP‘226 teaches Erlotinib hydrochloride and therefore must be taken to have raised a credible challenge to the validity of the IN‘774 patent.
 
• That Cipla argued that substitution of methyl with ethynyl would be obvious to a person skilled in the art in light of the closet prior art Gefitinib as claimed in the EP‘226. Cipla further argued that methyl and ethynyl are normally used interchangeably in chemical arts because they share common attributes and the Controller ought of have examined the EP‘226 while examining the IN‘774 patent. Roche, relying on TSM test, argued that a person of ordinary skill in the art would find no motivation at all to replace the methyl group at position 3 by an ethynyl group. The divisional bench reasoned that these arguments never considered in the Controller’s order in detailed manner which add to raise credible challenge on the validity of the IN‘774 patent.

• That Cipla argued that the Controller confused the concepts of inventive step, anticipation and obviousness and particularly confused the tests of inventiveness with obviousness which the divisional bench reasoned are not without merit. Cipla even raised creditability of articles referred to in the Controller’s order concerning enhanced efficacy as studies were conducted or sponsored by OSI Pharmaceuticals. The divisional bench agreed with Cipla that articles published in the journals were conducted by OSI and therefore lack creditability in proving enhanced efficacy of Erlotinib hydrochloride over the closet prior art. The divisional bench opined that the Controller misunderstanding over the tests of inventiveness and failure to notice that articles proving enhanced efficacy were sponsored by OSI raised credible challenge on the validity of the IN‘774 patent.

The divisional bench lastly weighed Roche argument whether the single judge erred in applying issue of “public interest” and “pricing” in rejecting the injunction. The divisional bench, citing Sec. 83 (e) and 83 (g), disagreed with Roche and opined that the element of “public interest” is not alien to the Indian patent law. The divisional bench even agreed with the single judge reasoning that the general public access to life saving drugs assumes greater significance than the grant of injunction.

The divisional bench reasoned that the public interest in greater public access to a life saving drug outweigh the pubic interest in granting an injunction to Roche as Cipla prima facie demonstrated that Roche do not hold patent for Tarceva tablet which is polymorph B of Erlotinib hydrochloride rather mixture of "polymorphs A+B" as claimed by the IN‘774 patent, and raised credible challenge on the validity of the IN‘774 patent.

While weighing the principles that should have taken into consideration in deciding grant of injunction in a suit for infringement of a patent, the divisional bench agreed with the single judge reasoning that the IN‘774 patent need scrutiny under Sec. 3 (d) and uphold the rejection order. The divisional bench further concluded finding no merit in any of the arguments made by Roche and even ordered Roche to pay Rs. 5 lakhs to Cipla.

The divisional bench decision was considerably difficult to pen-down not because of complexity but more because it reasoned contrary to basic principles and nuances of patent law and lacked even the simplest understanding of organic chemistry and polymorphism. In my seven years of patent practice, never came across such fictitious and disastrous observations as made by the divisional bench. On most counts, the divisional bench reasoning was not only ridiculous but found no substantive support in patent law.

On the point of “polymorphs A+B” what is unfortunate that the divisional bench agreed with Cipla’s fictitious argument by merely taking note of plain English excerpt from the US‘221 patent rather understanding it technically. Even the divisional bench’s understanding of patent law was grossly questionable. While deciding patent matters, be it simple or complex, what the divisional bench cannot ignore is the fact that patent is a techno-legal disclosure and cannot be given plain English reading and further patent speaks to ordinary persons familiar with/skilled in the art to which it relates to as contrasted with being written to be understood by an ordinary member of the public. 

Not only the divisional bench erred in properly understanding excerpt from the US‘221 patent but also ignored basic facts what an ordinary person familiar with the related art would have considered while reading such the excerpt.

• First and foremost, an ordinary person would have never confused over the term “mixture” used in the excerpt as a chemical compound can never a mixture in whatever context it was used.

• Second, chemical compounds are and have a fixed ratio of atoms (chemical elements) that are held together in a defined spatial arrangement by chemical bonds. These spatial arrangements define and distinguish the phenomena called polymorph.

• Third, a chemical compound exhibit different physical property due to change in the spatial arrangement of atoms but the compound will remain the same.

In view of above facts, an ordinary person would definitely have not understood what the divisional bench understood reading the term “mixture” and would have easily acknowledged the fact that the XRD data of Erlotinib hydrochloride taught in the US‘498 patent (US equivalent of the IN‘774 patent) showed traces (peaks) of both polymorph A and polymorph B. In other words, US ‘221 patent excerpt reveals to an ordinary person that Erlotinib hydrochloride prepared according to the US‘498 patent exhibit two different spatial arrangements. Nowhere an ordinary person may have concluded what the divisional bench agreed that the IN‘774 patent is for “polymorphs A+B”. In fact, the Controller of Patents while rejecting the IN‘507 application too opined that Erlotinib hydrochloride per se and polymorph B of Erlotinib hydrochloride are same and only differ in the spatial arrangement. In short, the divisional bench observation to restrict Erlotinib hydrochloride per se to “polymorphs A+B” and considering polymorph B of Erlotinib hydrochloride and Erlotinib hydrochloride per se altogether different compounds were technically flawed and without an element of merit.

What was even more alarming that while deciding appeal the divisional bench never bothered looking into claims of the IN‘774 patent not even once and repeatedly agreed or rather seem convinced with Cipla’s “polymorphs A+B” argument merely taking note of excerpt from the US‘221 patent and the fact that Tarceva product is polymorph B form of Erlotinib hydrochloride. Here the divisional bench either seem to have no idea about “claim construction” or deliberately ignored taking note of claims of the IN‘774 patent. 

It is a well-established principle that “claim construction” is a question of law to be decided by the court and the judge must construe the claims from the view of “one ordinary skill in the art at the time of invention.” But contrary to established principle, the divisional bench construed claims reading an excerpt from the US‘221 patent dated four years after the priority date of the IN‘774 patent and that also without looking into the claims. 

Further in patent infringement suits, the court decides whether the accused product infringes the claim but rather making decision on accused product Erlocip, the divisional bench focused on whether the Tarceva product is covered within the claims of the IN‘774 patent and yet again without looking into the claims. 

Another point where the divisional bench erred and opined contrary to nuances of patent law is when the divisional bench noted that the Controller of Patents failed to take into accounts facts pertaining to polymorph B while deciding the patentability of the IN‘774 patent which add credible challenge to the validity of the IN‘774 patent. In patent law, patentability of an invention is examined over the references published before the filing date not after the filing date. Sadly, the divisional bench seems to have even lacked basic understanding of patent law.

The divisional bench even erred on the point of “requirement of a full disclosure” when it noted that Roche failed to brought to the notice of the Controller of Patents that Erlotinib hydrochloride as claimed by the IN‘774 patent is mixture of polymorphs A and B. In patent law, “full disclosure” need to comply with the requirement of enablement and best mode. The IN‘774 patent is directed to Erlotinib hydrochloride per se and to comply the “requirement of full disclosure” the patentee must disclose the information sufficiently to enable a person to make the invention and the method by which patentee performed the invention. Sadly, here too the divisional bench seems to have messed with “full disclosure” requirement by adding polymorph issue. 

In short, the divisional bench not only erred in understanding the case from the view of “one ordinary skill in the art” but also made observations which stand contrary to nuances of patent law and established judicial principles. In next post will discuss the the single judge's decision on infringement issue.

In-depth analysis of Tarceva patent litigation

On September 07, 2012, the Delhi High Court in F. Hoffmann-La Roche et al v. Cipla Ltd. (No. 89/2008) ruled Indian Pat. No. 196774 jointly owned by OSI Pharmaceuticals and Pfizer for anticancer drug compound Erlotinib hydrochloride valid but not infringed by Cipla’s generic product Erlocip. In his painstakingly 275-page decision, the single judge, Justice Manmohan Singh opined that the generic product Erlocip manufactured and sold by Cipla contains polymorphic B variant of Erlotinib hydrochloride as active ingredient and hence do not infringed the compound per se claims of the IN‘774 patent. This case has avidly been followed by the life sciences industry with utmost interest and concern for guidance on how pharma patents will be adjudicated at judiciary level. It is likely that the reasoning of this decision may add yet another disappointment and further dismay for the pharmaceutical companies in enforcing their patents in India. Below, we discuss the decision in detail and meticulously examine the impact this case may have on patent protection for drug products in India.

A. Background of Tarceva patent

In March 1996, Pfizer and OSI Pharmaceuticals jointly filed the Indian Pat. App. No. 537/DEL/1996 for drug molecule Erlotinib hydrochloride and claimed priority from their corresponding US App. Sr. No. 08/413,300 dated March 30, 1995. The Application was originally filed with 27 claims but reduced to two claims during the examination proceedings and subsequently was placed in order for issuance on February 23, 2007.

On April 10, 2007, Natco Pharma filed an opposition against the issuance of the IN‘774 patent under Sec. 25 (1) of the Act (pre-grant opposition) on the grounds of obviousness and insufficient disclosure.

• Natco cited various European patent disclosures as prior art including Gefitinib patent disclosure but to substantiate the obviousness argument, Natco submitted a hypothetically created structure of 4-anilinoquinazoline nucleus with conditional substitution of four functional groups, namely, (1) 2-methoxy, (2) 2-methoxy ethoxy, (3) hydrogen and (4) 3-ethynyl on record to assert Erlotinib structure as obvious compound. However, Natco never provided any reasoning on what may have motivated the person skilled in the art to arrive with conditional substitution.

• Pfizer refuted Natco’s obviousness arguments and argued that the hypothetical structure is not found in any of the cited prior art disclosure and further emphasized inventiveness of substituting an “ethynyl group” at methane position of phenyl group. After hearing both Natco and Pfizer, the Controller found Natco’s obviousness argument unpersuasive and opined that mere structural similarity not sufficient to establish obviousness.

• Interestingly, Natco’s insufficient disclosure was neither based on lack of enablement requirement nor lack of best method requirement but rather odd ground that the application lack XRD data of the claimed compound. Pfizer refuted Natco’s argument and argued that they disclosed all the required information for a person skilled in the art to understand the invention and further emphasized the object of the invention was to provide potent EGFR inhibitor and not to provide polymorphic form of Erlotinib hydrochloride. After hearing Pfizer and Natco, the Controller found Pfizer’s argument persuasive and opined that the applicant has described the invention sufficiently to be reproduced by a person skilled in the art.

• Natco further raised Sec. 3 (d) objection during the hearing but the Controller strongly opined that once the invention has been found inventive, the invention cannot be held un-patentable under Sec. 3 (d) of the Act. 

On July 04, 2007, the Controller decided the pre-grant opposition and issued the IN‘774 patent for Erlotinib hydrochloride.

In another related development, OSI Pharmaceuticals filed a separate Indian Pat. No. IN/PCT/2002/507/DEL for polymorphic B variant of Erlotinib hydrochloride on May 14, 2002. The IN‘507 application is equivalent of the US Pat. No. 6,900,221 currently listed with the Orange Book that, in short, indicates Roche marketed Tarceva tablets contains Erlotinib hydrochloride matching the spatial arrangement of polymorphic B variant.

In January 2008, Cipla filed a pre-grant opposition against the issuance of patent for polymorphic B variant of Erlotinib hydrochloride on the grounds of insufficient disclosure and un-patentable under Sec. 3 (d) of the Act and the IN‘507 application was subsequently rejected on December 15, 2008. In this post will not discuss the rejection order in detail but one point that is worth highlighting here (and will be used later in our analysis) that the rejection of the IN‘507 application was primarily reasoned on the Controller’s opinion that compound Erlotinib hydrochloride per se and polymorphic B variant as disclosed in the IN‘507 application are same and only differ in the spatial arrangement of the molecule.

B. Suit for injunctive relief
In January 2008, various print-media newspaper reported Cipla planning to launch generic product Erlocip and that prompted Roche to file a civil lawsuit in the Delhi High Court on January 19, 2008 seeking an ex-parte interim injunction against Cipla. In reply, Cipla filed counterclaim to challenge the validity of the IN‘774 patent.

• To contest validity, Cipla argued claims of the IN‘774 patent lacks inventive step over prior published European patents that disclose similar quinazoline nucleus structure (including Gefitinib) and un-patentable under Sec. 3 (d) as Erlotinib is a mere derivative of Gefitinib. Cipla further argued that attempt to patent Erlotinib is ever-greening and contrary to public policy and against the statutory language employed in Sec. 3 (d). 

• To strengthen the validity challenge, Cipla questioned the IPO decision to grant patent for Erlotinib hydrochloride and reasoned that the issuance of the IN‘774 patent contrasted earlier IPO decision to decline issuance of patent for Gefitinib compound which share same 4-anilinoquinazoline nucleus structure as that of Erlotinib hydrochloride.

• While responding to injunction motion, Cipla argued that the injunction not be issued on the grounds of non-working of the IN‘774 patent and public interest with respect to pricing and affordability. 

On March 19, 2008, the court refused to grant interim injunction to restrain Cipla from selling the generic product Erlocip relying on the principles laid down in the American Cyanamid Co. v. Ethicon Ltd. The court found that Cipla’s arguments raise credible challenge to the IN‘774 patent and refrained from conducting a mini trial at the interlocutory stage. However, while taking note of Cipla’s argument that Erlotinib is nothing but a mere derivative of Gefitinib and lacks inventive step, the court went down critical on the IPO for taking formalistic approach in applying TSM test to examine the non-obviousness and further criticized the Controller for non-application of mind with respect of efficacy requirement under Sec. 3 (d). Further, the court weighed the “balance of convenience” in favor of the public interest to have access of a life saving drug rather granting injunction to affirm patent during pendency of infringement action but instructed Cipla to maintain account of sales and annual sales statement (duly authenticated by chartered accountant) and to furnish an undertaking to pay damages in case the IN‘774 patent found to be valid. 

It is interesting to note that Cipla’s counsel never pleaded non-infringement of claims of the IN‘774 patent but tactfully and categorically brought reference of the US’221 patent during the pleadings to construct a mirage-like effect over the issue of Erlotinib hydrochloride. While referring to excerpt from the US‘221 patent Cipla’s counsel argued that claims of the IN‘774 patent is a mixture of polymorph A and B forms of Erlotinib hydrochloride and further argued that Tarceva product sold by Roche is polymorph B form of Erlotinib hydrochloride. In short, Cipla’s counsel persuaded the judge that Tarceva sold in India is covered by the IN‘507 application which is corresponding of the US‘221 patent. Though this argument was nowhere a deciding factor is refusing interim injunction but was a master-stroke and game-changer strategy by Cipla that worked later in deciding infringement issue.

Though this suit was focused on the issue of interim injunction and the judge, Justice Ravindra Bhat, was considerably diligent in deciding “balance of convenience” but lacked same level of diligence while deciding on Cipla’s credible challenge on the validity of the IN‘774 patent. Cipla’s validity challenge was constructed around the rationale that Erlotinib is mere derivative of Gefitinib and should not be issued patent since the IPO declined to grant patent for Gefitinib. Roche refuted Cipla’s arguments using more or less same arguments that were made during the pre-grant opposition proceedings and were found persuasive by the IPO to grant patent. Even the court agreed that Cipla neither brought new prior art document nor new argument that was not taken into account during the pre-grant opposition proceedings. Further, the court emphasized that at the stage of interlocutory injunction defendant need not prove actual invalidity but must put forth a substantial question of invalidity to show that the claims at issue are vulnerable. In short, the court opined that substantial question of invalidity requires less proof than the clear and convincing standard to show actual invalidity.

On the point of “substantial question of invalidity” what is perplexing that the court agreed with Cipla on the “mere derivative” argument without even weighing the correctness of derivative correlation between Erlotinib and Gefitinib. Merely because Cipla’s counsel raised “mere derivative” argument, the court castigated the IPO for granting the IN‘774 patent but how without even explaining “what is derivative” and “whether Erlotinib qualify as Gefitinib derivative” the learned judge concluded that the IPO overlooked “mere derivative” insight during the pre-grant opposition proceedings and patent issued is vulnerable. It seems to be that whole premise of “substantial question of invalidity” was merely and plainly influenced by Cipla’s “mere derivative” theory as there was neither any documentary evidence to support it nor the court weighed the derivative correlation. This decision is, unfortunately, creating a dangerous precedent where an unsubstantiated argument may qualify as “substantial question of invalidity”.

As expected, Roche filed an appeal against the decision on April 11, 2008. In next post will discuss the divisional bench decision on Roche appeal and the Delhi High Court decision on infringement issue.