Pfizer, Inc - [2004] APO 12

OFFICIAL NOTICE
DECISION OF A DEPUTY COMMISSIONER OF PATENTS
Patent: No. 602638 in the name of Pfizer, Inc

Title: Triazole Antifungal Agents

Action: Request under s.70 for an extension of term of the patent

Decision: Issued 4 June 2004
Abstract

Application to extend the term of a patent under s.70. A subsequent 'selection' patent had previously had its term extended on the basis of an ARTG registration of the substance voriconazole. The same ARTG registration was being relied upon to extend the term of the 'parent' patent. During consideration of the request, the examiner objected that the relevant substance was not 'in substance disclosed' in the specification.
· where there is a selection patent that specifically claims a substance, there is a fundamental inconsistency with an assertion that the particular substance is in substance disclosed in the parent patent.
· The resolution of this inconsistency is necessarily an evidentiary issue. In the context of a request for extension of the term of a patent (which is likely to occur some considerable time following the filing of the respective patents), the following are relevant:
·The selection patent will likely have been applied for and granted at a time closer to the relevant date for construing the disclosure of the earlier patent, and therefore has weight in construing the disclosure of the earlier patent by reason of its temporal proximity – especially in comparison with evidence of skilled persons as of ‘today’;
·A selection patent that specifically claims a substance constitutes rebuttable evidence that that substance was not in substance disclosed in the parent patent;
·In circumstances where the substance is neither specifically identified or exemplified in the parent patent, the burden of rebutting the evidence provided by the selection patent necessarily lies with the person arguing that the relevant substance is in substance disclosed in the parent patent.

· Evidence provided was insufficient to rebut the evidence as provided by the existence of the selection patent. Extension refused.
PATENTS ACT 1990
DECISION OF A DEPUTY COMMISSIONER OF PATENTS
Re:Patent No. 602638 by Pfizer, Inc, and a request under s.70 for an extension of term of the patent.

BACKGROUND

In September 2002 Pfizer Pty Limited obtained ARTG Registration in respect of the therapeutic good VFEND voriconazole. The compound voriconazole has the chemical name 2R,3S-2-(2,4-difluorophenyl)-3-(5-fluoropyrimidin-4-yl)-1-(1H-1,2,4-triazol-1-yl)-butan-2-ol.
On 27 Feb 2003 Pfizer, Inc requested an extension of term of patent 625188 on the basis of that ARTG registration. In that patent, voriconazole is specifically claimed in claim 12 – as the 4^th of six specific alternative compounds.
Patent 625188 is in the nature of what is commonly referred to as a selection patent. At its commencement it states:
Some of the compounds of the present invention are disclosed in a general sense in our European patent application No 89307920.2 (EP-A-0357241) equivalent to patent application No 38930/89 but none of them are specifically described or exemplified therein.
The term of this patent was extended on 24 October 2003.
On 27 Feb 2003 Pfizer, Inc also sought an extension of term of Patent 602638 – which was granted on the aforementioned application 38930/89. The examiner objected that the substance voriconazole was not in substance disclosed in the specification. This objection was maintained in several reports, resulting in the issue being argued at hearing before me. [At the hearing on 25 March 2004, the patentee was represented by Mr R Webb of counsel, assisted by Ms K Howard of counsel, and Ms S Irani and Mr J O’Connor (patent attorneys, Spruson & Ferguson, Sydney)]
Noting that the patentee has obtained an extension of the term of its patent that specifically covers voriconazole, the following (from submissions filed on 10 September 2003) provides an explanation for why an extension is being sought on 602638:
34. Given that the intention of the extension of term legislation is to provide an effective fifteen year patent term, then extension of the term of the patent 602638 until 24^th July 2014 will go to meeting the intention of the legislation and provide an effective 12 year patent term in respect of the claimed active compounds. Extension of the later selection patent 625188 will only provide an effective 13.5 year patent term in respect of the limited compounds of formula 1 and in particular specific compound voriconazole and formulations of the same and not in respect of any of the other active compounds per se which fall within the scope of claim 1 of the 602638 patent. It is therefore crucial that the base patent be granted an extension of term.
As noted by the patentee, the Australian legislation does not preclude the extension of term of more than one patent based on a single ARTG entry. Rather the Patents Act 1990 requires inclusion in the ARTG of a good containing a substance, and requires the following in respect of the content of the specification:
70(2)(a)one or more pharmaceutical substances per se must in substance be disclosed in the complete specification of the patent and in substance fall within the scope of the claim or claims of that specification;
In the present case, there is no question about what is on the ARTG. Nor is there any question that the relevant substance (voriconazole) in substance falls within the scope of the claims of the patent 602638. It is also accepted that voriconazole is not explicitly identified in the specification of 602638. The sole question is whether voriconazole is “in substance disclosed” in the specification. If it is in substance disclosed, the extension must be granted. If it isn’t in substance disclosed, the extension must be refused.
The essence of the applicant's argument is that
'The fact that the later patent explicitly claims voriconazole is irrelevant to the question of whether that substance is in substance disclosed in the present specification. Section 70(2)(a) requires a determination uninfluenced by the mere fact that a 'selection' patent may have been granted or extended. Nothing within the words of the section prevents two patents from being granted by reason of one registration of one pharmaceutical substance in the ARTG.'
The applicant also argues that the examiner has erroneously rejected the application on the basis that the requirements for 'in substance disclosed' were the same as an enabling disclosure for novelty purposes; and that the examiner has erroneously equated the phrase 'in substance disclosed' to a requirement for explicit disclosure.' Finally, they argue that the evidence of a Dr Stamford is sufficient to establish that the substance was in substance disclosed, and the fact of the subsequent selection patent is of no probative value in the face of Dr Stamford's evidence.
At the hearing, I indicated that in my mind a fundamental problem with the argument that the substance was in substance disclosed was that it would then be possible to claim that substance per se. Consequently the subsequent selection patent would be invalid by reason of anticipation – either as a result of such a claim if it were inserted, or otherwise as a result of the disclosure. And that such an outcome would be inconsistent with the grant of the selection patent (assuming its validity).
Thus a key issue for consideration is what evidence can be considered to establish whether that substance is 'in substance disclosed' in the specification in circumstances where that substance is not specifically identified or exemplified. In particular, is it relevant to have regard to a subsequent selection patent – which (of its very nature) was filed after the filing of the specification in question? Or are the considerations limited to expert evidence about the disclosure of the specification, excluding any consideration of a subsequent selection patent? To this end, it is necessary to review some aspects of the law.
‘IN SUBSTANCE DISCLOSED’ and FAIR BASIS
Central to the present matter is the question of whether voriconazole is in substance disclosed in the present specification – and therefore what is meant by “in substance disclosed”. In many of the reported cases on this issue, the argument is about whether the disclosure is broader than the express words used. In other cases the issue is about whether a specific option in a larger class was in substance disclosed. And in some cases (for example Merck v Sankyo 23 IPR 415) the issue is about what was disclosed when there are errors in the specification. From my reading, the outcome is usually heavily dependent upon the facts. Indeed, as was stated in Merck v Sankyo:
30. Whether a matter is in substance disclosed in the specification as lodged is a question of fact to be determined in each case. As Terrell rightly points out in para. 8.19, reference to other decisions is often not very helpful; c.f. F. Hoffman-La Roche and Co Aktiengesellschaft v Commissioner of Patents (1971) 123 CLR 529. See also Minerals Separation Limited v Potter's Sulphide Ore Treatment Limited (1909) 8 CLR 779 per O'Connor J. at 800-801; Ricketson The Law of Intellectual Property para. 47.26.
The term 'in substance disclosed' relates to the extent of a disclosure. This issue arises in various contexts in the patent system. The main contexts are:
·    fair basis                real and reasonably clear disclosure
·    priority                  ‘fairly based on matter disclosed’ => real and reasonably clear disclosure
·    amendments          in substance disclosed
·    anticipation clear and unmistakable directions; mechanical equivalents
Interestingly the issue being dealt with in these different scenarios is essentially the same – viz. how different can something be from the express disclosure and still meet the relevant requirement. The fact that the 'tests' are expressed using different words naturally leads to a supposition that the criteria are materially different in some way. However, it is clear that the Federal Court considers the test for 'in substance disclosed' to be essentially the same as the 'real and reasonably clear disclosure' test of fair basis. By way of illustration:
In Merck v Sankyo 23 IPR 415, Lockhart J stated:
29. To determine whether proposed additional matter is in substance disclosed in the specification as lodged or constitutes new matter, the courts have applied basically the same test as they apply to determine whether complete specifications are fairly based on provisional specifications or convention patents. The Court of Appeal of the United Kingdom recognised the application of this test of fair basing in Ethyl Corporation's Patent (1972) RPC 169 at 194; and, as that judgment demonstrates, the tendency of the courts has been to give a liberal interpretation to sections such as s. 31(1) of the Patents Act 1949 (UK) (the equivalent of s. 49(4) of the 1952 Act), so as to permit any fair amendment which has already been in substance disclosed: see Ethyl per Denning M.R. at 194, 195. See also Shell Refining and Marketing Co Limited's Patent (1960) RPC 35 and AMP Incorporated v Hellerman Limited (1962) RPC 55; Terrell on the Law of Patents, 13th ed., 1982, para. 8.18.
In Gambro v Fresnius 49 IPR 321, the Full Federal Court stated:
18 His Honour noted, and it was not disputed before us, that it is a well-settled approach that the requirements of s 102 should generally be given a liberal construction because there is a strong public interest that inventive genius should be encouraged: Ethyl Corporation's Patent [1972] RPC 169. The requirement of substantial disclosure is very similar to the requirement that a claim must be "fairly based" on matter disclosed in a specification: Wimmera Industrial Minerals Pty Ltd v RGC Mineral Sands Ltd (No 3) (1997) AIPC 91-366 at 39,781-2; and, on appeal, RGC Mineral Sands Pty Ltd v Wimmera Industrial Minerals Pty Ltd (1998) 89 FCR 458. There must be "a real and reasonably clear disclosure": Société des Usines Chimiques Rhône-Poulenc v Commissioner of Patents (1958) 100 CLR 5 at 11; Leonardis v Sartas No 1 Pty Ltd (1996) 67 FCR 126 at 142-143. See also F Hoffman-La Roche & Co AG v Commissioner of Patents (1971) 123 CLR 529. As to the principles of construction, see Decor Corporation Pty Ltd v Dart Industries Inc (1988) 13 IPR 385 at 400. (my emphasis added)
In ICI v Lubrizol 49 IPR 513, the Full Federal Court stated:
118 The qualification is this. There is much authority for the proposition that there is a close relationship between the test for fair basing and the question whether matter is in substance disclosed in a specification. It is unnecessary to consider whether it is appropriate to go so far as to say that the two tests are "virtually the same" (Ethyl Corporation's Patent [1972] RPC 169 at 195). For the proposition that there is a close relationship there is no need to do more than refer to the decisions of the Full Court of this Court in CCOM at 280-282; Leonardis at 137-143; and RGC Mineral Sands at 460, 461. It will, we should think, be a rare case indeed where a claim which claims matter in substance disclosed in the specification as filed is not, equally, fairly based on the matter described in the specification (and vice versa): a comparison of his Honour's reasoning on amendment and ours on fair basing perhaps shows, with respect, why that is so. (my emphasis added)
And in RGC v Wimmera 42 IPR 353 at 356, Burchett J. stated, in discussing 'in substance disclosed':
In my opinion, it would be wrong to adopt a restrictive approach to s.102. The true limitation upon amendments imposed by the language of the section is that which Fullagar J, in the closely related context of fair basing, in Société des Usines Chemiques Rhone-Poulenc, that there must have been "a real and reasonably clear disclosure".
It follows that if a substance falling within the scope of a claimed generic formula is in substance disclosed in a specification (as filed), that substance per se can be made the subject of a claim, with that claim almost inevitably being fairly based on the description. That is, if voriconazole is truly in substance disclosed in the present specification, prima facie voriconazole can now be made the subject of a claim that is necessarily fairly based.
FAIR BASIS and CHEMICAL FORMULAE
The present case centres on the extent of disclosure that arises from the use of a chemical formula to represent the class of chemical compounds covered by the patent. This issue was the subject of the High Court case of F. Hoffman-La Roche v Commissioner (1971) 123 CLR 529. This case is usually relied on to support the proposition that a claim to a single compound included in the class is fairly based – even though the particular compound is neither specifically identified nor exemplified.
However that case merits careful consideration. It concerned a question of whether certain claims in the application were entitled to the date of the first basic application as their priority date. So the central question was fair basis on the disclosure in priority document – whether there was a real and reasonably clear disclosure of the relevant compounds. And the situation arose because the compounds were not specifically identified. Since the criteria for ‘fair basis’ and ‘in substance disclosed’ are effectively the same, the decision gives direct guidance on the question of ‘in substance disclosed’ in the context of chemical formulae.
Gibbs J concluded that the claims were fairly based on the disclosure in the first basic application. However, in reaching this conclusion he made a number of significant caveats. Firstly, when commenting about May & Baker Ltd. v. Boots Pure Drug Co. Ltd. (1948) 65 RPC 255; (1948) 66 RPC 8; and (1950) 67 RPC 23, which had been relied on by the Commissioner, Gibbs J stated:
The clear ground of distinction between the facts of that case and the present is that there the original claim was for a large class, all whose members were said to have therapeutic value, whereas it was discovered that only some members of the class had that value. In this case there is no suggestion at the present stage that the claim in the first basic application as to the utility of the compounds is false and there is therefore no ground for the submission that there is a fresh inventive step involved in selecting out of the class those compounds that are of utility. (My emphasis added)
He similarly commented on a passage in Ralston v. Smith (1865) 11 HLC 223, at pp 252-253 (11 ER 1318, at p 1329) which was cited by Lord Morton in the May & Baker Case (1950) 67 RPC, at p 47:
These remarks might be appropriate to the present case if only the compounds mentioned in proposed claims 2, 3 and 4 had the pharmaceutical utility claimed but that is not the situation. In the example given by Lord Cranworth, a combination, for example, of A and B would have been bad, whereas in the present case it has not been alleged that any compound within the description in the first basic application is lacking in utility. In other words, the appellant in the present case does not by the complete specification attempt to correct what was an imperfect general description in the first basic application; although it specially claims some of the compounds within the first basic application it does not do so on the footing that they alone are of utility or that they are possessed of a special utility which the others lack. (My emphasis added)
Finally, and of critical importance to the present matter, Gibbs J states:
22. It seems to me that the lines of investigation suggested by Lloyd-Jacob J. lead to the conclusion that proposed claims 2, 3 and 4 in the complete specification are fairly based on matter disclosed in the first basic application. It is crucial, in my view, that it is not suggested that the compounds the subject of proposed claims 2, 3 and 4 have been selected because they have any special utility. If a basic application disclosed a large class of compounds, all of which were claimed to be of pharmaceutical utility, and it were found that the claim was false, in that only some of the compounds were useful, or it appeared that some of the compounds had a particular and peculiar value, there would be much to be said for the view that a claim limited to those compounds selected for their utility or special value would not be fairly based on matter disclosed in the basic application, at least if the basic application did not itself provide a guide to that selection, and a fresh inventive step were necessary to enable it to be made. Here however a large class of compounds is disclosed, and clearly disclosed, in the basic application, and proposed claims 2, 3 and 4 in the complete specification are for compounds forming part of that class, but not selected because they alone are useful or because they have utility greater than that of other members of the class. (My emphasis added)
I note that this last passage was recently cited with approval in Fina v Halliburton 59 IPR 91 @ paragraph 102. See also the decision of Beaumont J in Coopers Animal Health v Western Stock 11 IPR 20 @ 36.
For a claim to a compound to be fairly based, there needs to be a 'real and reasonably clear disclosure' of that compound (as that term has been interpreted in numerous cases). However, where a specification defines a chemical class formulaically, and:
· the particular compound has been selected because the other members of the class did not have the indicated utility of that class; or
· the particular compound has been selected because of some special utility;
a claim to that compound cannot be fairly based.
'IN SUBSTANCE DISCLOSED' and SELECTION PATENTS
The logical consequence of the above, in the context of an individual member of a formulaically defined class (which member has been neither specifically identified or exemplified) is:
· if it is true that the member is 'in substance disclosed' in the specification, a claim to that member per se would be fairly based; but
· if that member had been 'selected' because of some special utility, a claim to that member per se would NOT be fairly based.
That is, the member is NOT 'in substance disclosed' in the specification in circumstances where there is some special (inventive) utility associated with that substance that is not identified in the specification.

Selection patents are an express situation where compounds are selected for a special utility. The rules for the grant of a selection patent were set out in I.G. Farbenindustrie (1930) 47 RPC 289 at pages 322/3 as
a)the selection must be based on some substantial advantage gained or some substantial disadvantage avoided;
b)the whole of the selected members must possess the advantage in question; and
c)the selection must be in respect of a quality of a special character which may fairly be said to be peculiar to the selected group.
Consequently, the existence of a subsequent selection patent in respect of that particular member (or a subclass containing that member), is prima facie evidence that the member has a special utility not identified in the earlier patent. And this evidence may arise from assertions made in the specification, as well as the fact of its grant following examination having regard to the issues of anticipation that might otherwise exist.
Accordingly, I am of the view that in circumstances where there is a selection patent that specifically claims a substance, there is a fundamental inconsistency with an assertion that the particular substance is in substance disclosed in the parent patent.
However the situation is one of inconsistency, not of uncontroverted fact. It might ordinarily be presumed that the assertions made in the specification of a selection patent about the surprising advantages to be found in the selected members were founded in fact. Or that the patent was properly granted having regard to the otherwise anticipatory effects of the parent patent. But it might also be the case that the selection patent contains false assertions of fact, or was invalidly granted. The resolution of this is necessarily an evidentiary issue, and in the context of a request for extension of the term of a patent (which is likely to occur some considerable time following the filing of the respective patents) I think the following are relevant:
·    The selection patent will likely have been applied for and granted at a time closer to the relevant date for construing the disclosure of the earlier patent, and therefore has weight in construing the disclosure of the earlier patent by reason of its temporal proximity – especially in comparison with evidence of skilled persons as of ‘today’;
·    A selection patent that specifically claims a substance constitutes rebuttable evidence that that substance was not in substance disclosed in the parent patent;
·    In circumstances where the substance is neither specifically identified or exemplified in the parent patent, the burden of rebutting the evidence provided by the selection patent necessarily lies with the person arguing that the relevant substance is in substance disclosed in the parent patent.
It is my view that where a substance said to have been in substance disclosed in the parent patent (but is not specifically identified or exemplified) is specifically claimed in a selection patent, the patentee of the parent will need to provide clear and convincing evidence to support their contention that it is in fact ‘in substance disclosed’ in that specification. The issue is necessarily a question of fact, and ‘argumentative’ evidence is unlikely to be of much assistance. Furthermore, assuming the selection patent was filed under the 1990 Act such that ‘whole of contents novelty’ is available (or if under the 1952 Act, the parent application had been published before the relevant priority date of the selection patent – or a relevant prior claim situation exists), it is noteworthy that such evidence (if credible) may impugn the validity of the selection patent by reason of anticipation.
As previously noted, the patentee has argued that I should not have any regard to the later selection patent when interpreting the disclosure of the present patent; that the present patent should stand on its own. I disagree. It might be an appropriate approach when the matter of disclosure is simply one of grammatical construction – but it cannot be the case when evidence is required to establish disclosure. And in the context of what evidence might be considered, I think it is entirely artificial to suggest that the Commissioner should ignore the existence of any material other than the patent specification itself, and any other material that the patentee chooses to put before her. The issue of disclosure (in the absence of explicit disclosure) is necessarily an evidentiary issue. Selection patents by their very nature involve a special relationship with the parent patent, as set out in the IG Farbenindustrie rules. This relationship necessarily means that there is forensic value arising from the existence and grant of that patent. It may be that the grant of the selection patent was in error. It may be that other evidence is of much greater weight, or is conclusive. But the selection patent will usually constitute relevant evidence as of the date of filing the selection patent, of what the applicant considered was disclosed in the parent. Accordingly, I consider it is appropriate for the Commissioner to have regard to the disclosure of a relevant selection patent as evidence in proceedings under s.70.
THE EVIDENCE
In the present case, the extension of term is based on the substance voriconazole. This substance is claimed explicitly in claim 12 of the selection patent, as the 4^th of six specific alternatives. The specification of the selection patent includes the statement (4^th paragraph of page 1A):
It has now been discovered that the compounds of the present invention have a surprisingly high level of fungal activity, in particular against Aspergillus spp. fungi, which is mainly attributable to their unexpectedly good pharmacokinetic properties which result in longer half-lives (t^½ values).
Additionally, I note that the parent patent was cited during examination by way of a prior claiming objection, and the 'surprising' aspect referred to in the above paragraph was reiterated in rebuttal – clearly indicating that the selection criteria were relevant to the grant of this patent. Either of these alone is sufficient to make it prima facie clear that voriconazole (and the other substances in that claim) has a special utility not identified in the present patent. Following from F. Hoffman-La Roche v Commissioner, voriconazole cannot have been in substance disclosed in the present specification – assuming the correctness of the assertions in the specification of the selection patent, of the assertions made during prosecution, and otherwise its validity given the existence of the parent patent. That is, patent 625188 provides strong (although rebuttable) evidence that voriconazole is not in substance disclosed in present patent 602638.
The patentee has provided two declarations in support of its contention that voriconazole is in substance disclosed in the present specification. The most relevant of these is that by N. Stamford. He states that he has been
'asked to give my opinion as to whether Australian patent 602638' includes the molecule called voriconazole.'
He goes on to state:
‘Given the synthesis described in Example 3, it would not be difficult to prepare a fluorine substituted derivative’ (that is, voriconazole)
and
‘In conclusion, amongst the structures included in Formula I in the patent, I can clearly identify voriconazole.’
The patentee asserts that the declaration by Stamford establishes that voriconazole is in substance disclosed in the specification. I disagree. The declarant compares the compound against the various requirements of the structural formula, making observations that the particular substituents present in voriconazol match the requirements of the chemical formula. The declarant also notes that the compound of example 3 is closest to voriconazol, with
'the only difference being that it does not have a fluorine substituent in position 5 on the pyrimidinyl ring. Given the synthesis described in Example 3, it would not be difficult to prepare a fluorine substituted derivative.'
On a proper consideration, all this declaration establishes is that voriconazole is one of the chemical entities covered by the chemical formula, and that the description was sufficient to enable its manufacture if such was desired. It fails to address the question of whether the substance was 'in substance disclosed' in the specification.
The second declaration is by a D Ellis. It provides some background on antifungals. However it does not provide any relevant evidence of the disclosure of patent 602638.
Accordingly the situation is:
·    The patentee asserts that voriconazole is in substance disclosed in the present patent 602638;
·    The existence of the selection patent 625188 (including statements in the specification, and made during prosecution) constitutes clear evidence that voriconazole is not in substance disclosed in 602638;
·    The evidence of Stamford goes to whether voriconazole falls within the scope of the generic disclosure and whether there is an enabling disclosure – not whether it was 'in substance disclosed'. The evidence of Stamford and of Ellis is manifestly insufficient to rebut the evidence as provided by the existence of the selection patent.
On this basis, I am not satisfied that voriconazole is in substance disclosed in patent 602638. Accordingly the criterion of s.70(2)(a) are not met, and the request for an extension of term of this patent must be refused.
D Herald
Deputy Commissioner of Patents
04 June 2004
Patent attorneys for the patentee : Spruson & Ferguson, Sydney