As promised in our last post, here is a more detailed analysis of the Supreme Court’s judgement in the Novartis case rejecting the patent application for the beta-crystalline form of imatinib mesylate (known as Glivec – “claimed invention”).
But first, here are some links to the coverage of the case by some of the top blogs in the U.S. and the U.K.: Patently O which featured two posts, one by Professor Dennis Crouch and the second post on the same blog by Prof. Srividya Raghavan. Here’s a link to the post on IPKat by Stefano Barazza. The Indian Express and DNA published op-eds by Shamnad. A number of other newspapers published an op-ed by Mr. Anand Grover, who appeared as the Senior Counsel for the Cancer Patient Aid Association in this case. Finally, here is an editorial of the New York Times on the judgment and its verdict. The NYT, rightly classifies the judgement as a ‘limited precedent’.
I have covered the basic timeline of the case in my earlier post over here and in the interests of brevity, I’ll jump directly to the main points in contention.
(i) What was the invention in question?
There appears to be confusion in some quarters, as to the exact invention in question. Just to clarify, the patent specification in question, can be accessed over here (Sai had put it up some months ago) and Claim 1 on page 23 of the specification informs us that subject of the invention was only the beta-crystalline form of imatinib mesylate.
The subject of the invention was therefore only the beta-crystalline form of imatnib mesylate and not imatinib mesylate itself.
(ii) Was the claimed invention found to be anticipated or lacking in novelty?
A claimed invention is said to be ‘anticipated’, or lacking in novelty, in patent law when the entire invention is disclosed, as is, in a single piece of prior art. The policy reason for this is simple: patents are granted for something new and if the claimed invention has already been described there is no reason to give it a patent. Section 64(1)(e) of the Patent Act allows for a patent to be revoked in case it was anticipated by a piece of prior art.
In the present case, the sequence of events with crucial prior art was as follows:
(i) The Zimmerman patent in ’93, which disclosed the imatinib free base – this patent was not granted in India because India did not provide for pharmaceutical patents prior to 1995;
(ii) A publication in 1996, which referenced imatinib mesylate;
(iii) The 1998 patent application before the Indian Patent Office which claimed the beta-crystalline form of imatinib mesylate.
For a finding of ‘anticipation’, the Court would have to be provided with evidence that the ‘beta-crystalline’ form of imatinib mesylate had already been in use in India or published in single document, anywhere in the world.
The Supreme Court however does not find any evidence that the beta-crystalline form of imatinib mesylate is anticipated. In fact in para 158, it states “This leaves us with the beta crystal form of Imatinib Mesylate, which, for the sake of argument, may be accepted to be new, in the sense that it is not known from the Zimmermann patent.”
The Court however spend a significant portion of the judgement explaining that imatinib mesylate is anticipated. This analysis was a necessary precursor to the Section 3(d) analysis because the provision requires the beta-crystalline form of imatinib mesylate to be compared with a “known substance” from which it is derived.
Novartis had tried to argue that the beta-crystalline form of imatinib mesylate should be compared with imatinib. However, the generics wanted the comparison to be with imatinib mesylate, since that would make it more difficult for Novartis to prove a significant increase in efficacy.
In order to establish this argument, the generics had to first establish that imatinib mesylate was “known” in the period intervening the ’93 Zimmerman patent for imatinib and the beta-crystalline form of imatinib mesylate.
There Court accepted the argument that there was significant prior art, including a judgement from the U.S. Board of Patent Appeals, which reportedly held that imatinib mesylate was directly anticipated from the ’93 patent.
The court thus came to the conclusion that at least imatinib mesylate was anticipated by the prior art documents.
(iii) Section 3(d) Analysis: The court then moves to evaluate the patentability of the beta-crystalline form of “imatinib mesylate” as per the requirements of Section 3(d).
(a) Does Section 3(d) even apply?
The first argument put forth by Novartis, was that Section 3(d) would not even apply, since it required a comparison of the claimed invention with a ‘known substance’ having known ‘efficacy’ and that neither imatinib nor imatinib mesylate had any known efficacy.
The Supreme Court dismissed this argument on the following grounds:
(i) That the Zimmerman patent of 1993, had identified the tumour treating potential of the imatinib free base and its derivatives, such as imatinib mesylate.
(ii) That prior art documents like the article published in Cancer Research clearly identified the in-vivo experiments carried out with imatinib mesylate.
(b) What is the “known substance” for the purpose of Section 3(d)?
As explained above, Section 3(d) requires the claimed invention to be more efficacious than the ‘known substance’ from which the claimed invention was derived.
In this case, Novartis was keen to have the imatinib free base identified as the ‘known substance’ instead of imatinib mesylate since it would be easier to prove greater efficacy vis-à-vis the imatinib free base. In support of such an argument, Novartis tried to reason that the beta crystalline form of imatinib mesylate was derived directly from imatinib and not imatinib mesylate.
The following is the relevant extract from para 170 of the judgment:
“The whole case of the appellant, as made out in the subject application and the affidavits, is that the subject product, the beta crystalline form of Imatinib Mesylate, is derived from Imatinib, and that the substance immediately preceding the beta crystalline form is not Imatinib Mesylate but Imatinib in free base form. This position is sought to be canvassed in the subject application and the affidavits on the premise that the Zimmermann patent ended at Imatinib in free base and did not go beyond to Imatinib Mesylate.”
The Supreme Court shot down this argument on the following grounds:
(i) That imatinib mesylate was anticipated by prior art as already discussed above and that it existed before the claimed invention and was hence a known substance;
(ii) That Novartis had allegedly described imatinib mesylate as a necessary step to produce the beta crystalline form of the drug from imatinib. Following is the relevant extract from para 170:
“Not only is this premise unfounded as shown earlier, but the appellant itself appears to take a somewhat different stand, as before this Court it was contended that the subject product, in terms of invention, is two stages removed from Imatinib in free base, and the substance immediately preceding the subject product is Imatinib Mesylate (non-crystalline).”
Therefore for the purposes of this case, imatinib mesylate was presumed to be the known substance. The court then proceeds to define the scope of efficacy, for the purpose of comparing the known substance with the claimed invention.
(c) What is the meaning of “efficacy” in Section 3(d)?
The most important limb of the judgement is the interpretation of ‘efficacy’ in Section 3(d). While the generics, the patient groups and Shamnad were arguing for interpreting efficacy as only therapeutic efficacy, Novartis was arguing for a broader interpretation which would include other beneficial properties such as increased stability etc., even though such properties would not lead to an increase in efficacy.
The Supreme Court however makes it crystal clear that ‘efficacy’ in Section 3(d) means only therapeutic efficacy.
In para 180, the Court states “What is evident, therefore, is that not all advantageous or beneficial properties are relevant, but only such properties that directly relate to efficacy, which in case of medicine, as seen above, is its therapeutic efficacy.”
The Court therefore clearly rejects the arguments by Novartis for a more broad based interpretation of efficacy which would have included even non-therapeutic efficacy such as properties which contribute to better storage and ease of administration.
The Court bases its conclusion on the manner in which the word ‘efficacy’ has been used in main text of Section 3(d) and the explanation. Given the context in which the phrase ‘efficacy’ is used, the court reasons “that not all advantageous or beneficial properties are relevant, but only such properties that directly relate to efficacy, which in case of medicine, as seen above, is its therapeutic efficacy.” This was quite similar to the arguments put forth by Shamnad in his intervention application, which can be accessed over here.
On the exact scope of therapeutic efficacy itself, as discussed in the earlier post, the Court refused to rule on the differing interpretations put forth by Mr. Grover and Shamnad. The scope of therapeutic efficacy is therefore an open issue.
(d) Does an increase in bioavailability qualify as increase in therapeutic efficacy under Section 3(d)?
One of the main points of contention, was whether or not the 30% increase in bioavailability of the beta-crystalline form of imatinib mesylate when compared with imatinib, would qualify as an increase in therapeutic ‘efficacy’ as required by Supreme Court’s interpretation of Section 3(d)?
Bioavailability is basically the increased ability of the drug to dissolve into the bloodstream of the patient.
On this point, the Court, depending on an authority provided by Shamnad, ruled that such an increase in bioavailability can qualify for protection under 3(d), if evidence is provided to establish that such an increase leads to greater therapeutic efficacy. Thus patent applicants will have to establish a link between the increased bioavailability to the increase in therapeutic efficacy of a drug.
(e) Was the claimed invention by Novartis more efficacious than the substance that it was derived from?
In the final limb, the Court had to compare the efficacy of the claimed invention with the known substance.
Very confusingly, the Court compares the claimed invention with not only imatinib mesylate which was identified, earlier in the judgment, as the ‘known substance’ for the purposes of Section 3(d) but also with the imatinib free base, despite having earlier rejected Novartis’s plea to consider the imatinib free base as the ‘known substance’ for the purpose of a Section 3(d) analysis.
Both comparisons are as described below:
(1) Comparison with imatinib mesylate: When compared with imatinib mesylate flow properties, better thermodynamic stability and lower hygroscopicity.
The Court then comes to the conclusion that none of the three properties identified above would qualify under a Section 3(d) analysis since they would not contribute to an increased therapeutic efficacy. The relevant extract of the judgment is as follows:
“173. The aforesaid properties, (“physical attributes”according to Manley), would give the subject product improved processability and better and longer storability but, as we shall see presently, on the basis of those properties alone, the beta crystalline form of Imatinib Mesylate certainly cannot be said to possess enhanced efficacy over Imatinib Mesylate, the known substance immediately preceding it, within the meaning of section 3(d) of the Act.”
(2) The comparison with the imatinib free base: The court then concludes that, even if it were to indulge Novartis by comparing the imatinib free base and beta crystalline form of imatinib, it could not possibly rule on such a comparison because Novartis did not provide suitable evidence to this effect.
The judgement holds in para 189 “In this case, there is absolutely nothing on this score apart from the adroit submissions of the counsel. No material has been offered to indicate that the beta crystalline form of Imatinib Mesylate will produce an enhanced or superior efficacy (therapeutic) on molecular basis than what could be achieved with Imatinib free base in vivo animal model.”
I think it is safe to conclude that the judgement was well reasoned, for the most part. It would have helped if the judges used some sub-headings in the judgement. Like I’ve always said, the hallmark of good judges like Justice S.B. Sinha or Justice Ravindra Bhat is the use of sub-headings.
It is also safe to conclude that the judgement will act as only a limited precedent because it was very fact specific. Of course, the debate on the interpretation of efficacy is now a settled topic since the Supreme Court has unequivocally interpreted the term to mean only ‘therapeutic efficacy’. The Court has however left open the question of how exactly to interpret ‘therapeutic efficacy’.
Some aspects towards the end of the judgement appear to be poorly reasoned, such as the point on whether or not Novartis was selling imatinib mesylate or the beta-crystalline form of imatinib mesylate. The Court concludes that Novartis was selling the former and not the latter and comes to this conclusion on the basis of the packaging of Glivec. This is a shaky conclusion since, the Drugs and Cosmetics Act specifies certain norms on labelling and the Court should not have come to any such conclusion without referencing the law on the point. More importantly, it makes no sense to raise such issues in the context of a patent validity case. If anything the Court should have referred the matter to the DCGI for relevant action.