Overbreadth Argument Rejected as Being Akin to the Promise Doctrine

Eli Lilly Canada Inc v Apotex Inc 2020 FC 814 St-Louis J

2,492,540 / tadalafil / CIALIS ADCIRCA

In this case, Lilly asserted that the processes used by Apotex to make tadalafil infringed Lilly’s 540 process patent. In addition to non-infringement, Apotex argued that the 540 patent was invalid for anticipation, obviousness, lack of utility and overbreadth. St-Louis J held that while the 540 patent would be infringed if it were valid, it was invalid for anticipation and obviousness, though the attacks based on lack of utility and overbreadth failed [204]. So far as I can tell, the holdings of invalidity on the basis of anticipation and obviousness turned on the facts and not on any point of law, though I have to say that the facts were quite complex and it is possible that I missed an issue.

There were a few other points of interest: (1) St-Louis J rejected an overbreadth attack as being akin to the promise doctrine; (2) there was a discussion of the presumption relating to process patents; and (3) there was a discussion of obvious errors in a claim, which related to the utility argument.

Overbreadth

Apotex argued that the 540 patent was overbroad on the basis that a particular (redacted) solvent has been tested and found not to work, but the disclosure nonetheless stated that it was useful [350].

St-Louis J rejected this argument, saying:

[356] The way that Apotex articulated the allegation of overbreadth in this case appears indeed very akin to the promise doctrine, abolished in AstraZeneca SCC. What Apotex really asks the Court to do is to parse the disclosure, conclude that [redacted solvent] promises to be useful for the PSR, import [redacted solvent] into Claim 1 in the absence of any ambiguity, and strike Claim 1 as a result.

She noted that the SCC in AstraZeneca SCC warned against this, and stated that “As such, the doctrine of overbreadth should not be applied in the manner suggested by Apotex, akin to the promise doctrine” [358].

St-Louis J’s comments strike me as entirely sound (with the caveat that her description of Apotex’s argument at [350] is very brief, so that can’t assess for myself her characterization of it at [356]). I have a paper on “Overbreadth in Canadian Patent Law” forthcoming in the IPJ, in which I argue that the approach to overbreadth adopted in Amfac (1986), 12 CPR (3d) 193 (FCA) “if widely adopted, risks invalidating patents for inventions which are new, useful and non-obvious, on the basis of an arbitrary parsing of the disclosure, in a manner reminiscent of the promise doctrine.” A draft version is available on SSRN (note that this draft was updated on 10 June from the first SSRN version). In that paper, I also show that overbreadth is almost always redundant, as merely restating a statutory ground of invalidity (most commonly lack of sound prediction of utility). In this case, if I understand Apotex’s argument correctly, it would appear that the preferable statutory basis for the attack would have been s 53(1); if St-Louis J had allowed the attack to go forward on the basis of overbreadth, this would have side-stepped the statutory requirements and related case law requiring wilfulness and materiality.

Presumption regarding a process patent

The 540 patent is a process patent, and Lilly argued that the burden should be reversed, so that Apotex would have to prove non-infringement on the basis of s 55.1, which provides that when the claim is to a process for obtaining a “new product” it should be presumed that the product was produced by the patented process. Lilly wished to argue that “new” meant that the product had not been sold on the market before, while Apotex argued it mean “new” in the sense of being previously known, whether or not it had received marketing authorization: [37], [38]. The prior caselaw at the FC level is against Lilly, which therefore did not press the point, but merely sought to preserve its rights on appeal [41].

Lilly also argued that the burden should be reversed on the basis of the common law rule that is more or less to the effect that “when the subject-matter of the allegation lies particularly within the knowledge of one of the parties, that party must prove it, whether it be an affirmative or negative character” [42]. The exact nature of the common law presumption is unsettled. In Cefaclor, 2009 FC 991, [221], Gauthier J held that the presumption would have applied “given the particular circumstances of this case,” if “Lilly had taken reasonable steps to obtain this information.” St-Louis J relied on this to hold that the presumption did not apply because the evidence did not allow her to conclude “that Apotex did not diligently seek to provide the requested process documents, nor that Lilly diligently sought further information from Apotex” [45]. This is reasonable enough as an application of Gauthier J’s holding to the facts of this case, though the caveat in Cefaclor regarding “the particular circumstances of this case,” suggests that the holding might be a fairly narrow one. It will be interesting to see how this line of reasoning develops in future cases.

Obvious Error in the Claim

Claim 12 had an obvious error. It read:

12. A method of preparing [tadalafi] comprising the steps of: . . .

(b) reacting [C] with [D] to provide [E];

            (c) reacting the product of step (b) with [F] and [G] to provide [E]; . . .

The product of step (b) is E, so reacting it with F and G would clearly not provide E. In fact, step (b) would not produce E, but rather a different compound, not otherwise mentioned in the claim [197]. The claim was otherwise correct, so that carrying out step (c) on the product of step (b) would indeed produce E.

It appears Apotex acknowledged that the error was obvious [343], and St-Louis J found on the facts that indeed a skilled person “would understand Claim 12 to bear [sic] a mistake, and would make tadalafil by following the sequence of actions” [349]. Nonetheless, Lilly did not ask the court to correct the error [197], [347]; instead “they are asking the Court to simply accept the evidence of the experts providing how a skilled person would read Claim 12c. Essentially, Lilly argue that the skilled person understands the error and accordingly understands the scope of the claim, which accords with the purposive approach construction” [197]; and see [347].

There is ample authority holding that an obvious error this type will not affect the construction of the claims: see eg Procter & Gamble (1979), 42 CPR (2d) 33, 36-37 (FCA); Cefaclor 2009 FC 991, [159]; Lovastatin 2010 FC 1265, [99]. Varco 2013 FC 750; Azithromycin 2005 FC 1421, [36]. The notion that the claim is not being “corrected” but merely read as a skilled person would read it is a technical distinction which is somewhat difficult to grasp.

In Procter & Gamble at 37 the FCA simply held that “the claims should be interpreted in the only way that makes sense”; in Cefaclor at [159] Gauthier J stated that a skilled person would “understand” the claim to apply as it should have been written; in Lovastatin at [99], Snider J held that the error “would not change the meaning ascribed to the phrase by the skilled addressee”; in Azithromycin at [36] Mosely J stated that the incorrect word in the claim is a typographical error “and is of no moment.” These statements are all a bit ambiguous as to whether the error was being “corrected” or rather that the claim was being read as a skilled person would read it. Phelan J in Varco at [336], on the other hand, expressly held that the error would be corrected: “[the plaintiff’s expert] opined that a Skilled Person would see the error and make the necessary corrections; and would not be confused or misled. [The defendant’s expert] all but admitted the same. Even a judge hearing this case could see the error and make the correction” (my emphasis).

In any event, a person untrained the law would say that an obvious error in the claim will be corrected. Even for a lawyer, that’s the easiest way to remember the rule. However one phrases it, the rule is very well established, with St-Louis J’s decision adding to the already ample authority.

 

Routine Correction of Inventorship

Alfasigma SPA v Canada (Attorney General) 2020 FC 561 Grammond J

            2,538,546

This was a routine correction of inventorship under s 52. One inventor has been named on the Italian priority application, but his name had inadvertently been left off the PCT application [3]. The application was supported by affidavits of all the other inventors [4]. No interesting issues were raised and the application for correction was granted [5].

Correcting Inventorship Post-Grant

Inguran LLC dba STgenetics v. Canada (Commissioner of Patents) 2020 FC 338 Brown J

2,518,882 / 2,752,218 / 2,752,244 / 2,752,247

This decision concerned an uncontested application to remove six named inventors from several granted patents. Brown J followed Qualcomm 2016 FC 499 (here) in holding that an affidavit from remaining inventors to the effect that they are the sole inventors is not required, so long it is adequately established on the other evidence that the named inventors to be removed are not in fact inventors. The case also illustrates that the power of the FC to correct inventorship is post-grant pursuant to s 52 is broader than the power of the Commissioner to correct inventorship pre-grant pursuant to s 31.

In this case, the original application had twelve named inventors. After a unity of invention objection by CIPO, the application was divided into divisional applications relating separately to electro-mechanical subject-matter and biological subject-matter. All twelve inventors were left on the divisionals, even though the named inventors sought to be removed in this application had not contributed to electro-mechanical subject-matter claimed in the patents at issue. This application sought to remove those non-inventors.

Before a patent is granted, inventors can be removed by the Commissioner pursuant to s 31(3) and added by the Commissioner pursuant to s 31(4). The Commissioner does not have the authority to add or remove inventors after grant: Micromass 2006 FC 117. (The Commissioner can correct some essentially clerical errors under the Patent Rules, including a spelling error in the name of an inventor pursuant to Rule 109 (this was previously authorized by s 8, now repealed).)

Post-grant, grant, inventors may be added or removed on application to the FC pursuant to s 52, which gives the Court broad authority to “order that any entry in the records of the Patent Office relating to the title to a patent be varied or expunged.” A number of decisions by the FC have invoked the test under s 31 in a decision pursuant to s 52, notwithstanding that on its face s 31 applies only to the Commissioner in respect of applications (see here). Presumably this is because s 31 provides more specific guidance regarding changes in inventorship specifically.

There is a bit of a concern regarding this practice, as s 31 is more restrictive than the broad powers granted to the Court by s 52. In particular, s 31(3) regarding removal of non-inventors (ie named inventors who in fact did not contribute to the development of the claimed inventions), specifies that the application may be carried on by the remaining inventors “on satisfying the Commissioner by affidavit” that they are the true inventors. S 31(4), regarding addition, provides that inventors may be added on satisfying the Commissioner that (i) they should be so joined, and (ii) that the omission of the further applicant or applicants had been by inadvertence or mistake and (iii) was not for the purpose of delay. The question is whether these—in particular (ii) and (iii) are all required under s 52 as well.

Brown J agreed with the submission that in an application under s 52 the Court “may apply” the tests set out for the Commissioner of Patents in subsections 31(3) and 31(4) [35], citing various cases. I’d quibble a bit on this point, at least to the extent that “apply” suggests that the test under s 31 is also the strict test under s 52. Section 31 was certainly considered in the cases cited by Brown J, but I’m not sure it was actually adopted as the test under s 52, rather than as identifying relevant considerations. While the application to correct inventorship under s 52 will certainly be granted when the test under s 31 is satisfied, the question is whether it is necessary to satisfy the test under s 31 to correct inventorship under s 52. Brown J’s decision suggests that the answer is no.

First, Brown J noted at [37] that in Imperial Oil 2015 FC 1218 Leblanc J set out two criteria for the removal of named inventors set out by subsection 31(3):

[15] Subsection 31(3) of the Act sets out two criteria for the removal of named inventors:

i. does it appear that one or more of the named inventors have no part in the invention?; and

ii. has an affidavit been provided to satisfy the Court that the remaining inventors are the sole inventors?

However, in this case, two of the six remaining true inventors could not be located [29] and so they did not provide affidavits. There was plentiful evidence from the other named inventors, including both the true inventors and the non-inventors, that the non-inventors were indeed not inventors, and Brown J followed Qualcomm 2016 FC 499 (here) in holding this was a housekeeping requirement that is “not relevant when an issued patent is being considered by the Court under section 52 of the Act” (Qualcomm [10], quoted at [39]).

Thus, whether or not affidavits from all remaining applicants are required under s 31(3), there is no such requirement under s 52. This indicates that the test under s 52 is not the same as the test under s 31(3).

This makes sense functionally. It may be much more difficult to find the remaining inventors post-grant, as is illustrated on the facts of this case. Further, while it may make sense for the Commissioner to want to be satisfied that the remaining applicants are in fact the inventors before the patent is granted to them, it is not clear why the FC would need to be satisfied that the remaining inventors are in fact inventors, given that the Commissioner was already satisfied of this fact on granting the patent. Surely, post-grant, that can be presumed. The real question post-grant is simply whether the inventors sought to be removed were not in fact inventors, and it would seem that affidavits from those non-inventors to that effect would normally be sufficient evidence. I suppose the point will really be tested if one named inventor is to be removed, on the basis of satisfactory evidence that the named inventor not in fact an inventor, in circumstances in which none of the remaining named inventors can be located. It seems to me that it would be clearly appropriate for the FC to order the non-inventor to be removed in such circumstances.

Routine Correction of Inventorship

General Transport Equipment Pty Ltd v Canada (Attorney General) 2019 FC 309 McDonald J
            2,639,371

In the third decision released this week that raises no interesting issues of law, McDonald J granted a routine uncontested application, pursuant to s 52, to amend the records to add a co-inventor to the 371 patent.

Is the Test for Correcting Inventorship Errors under S 52 the Same as under S 31?

Pharma Inc. v. Canada (Commissioner of Patents) 2019 FC 208 Pentney J
            2,503,570

This decision concerned an uncontested application to add the name of a co-inventor who had been inadvertently omitted from the original application [1], [2]. The leading case on this issue is Micromass UK Ltd v Canada (Commissioner of Patents), 2006 FC 117 [10], which held that an application to the Court is required, as the Commissioner cannot make corrections related to inventorship under s 8 of the Act, which only gives the Commissioner authority to correct “clerical errors” [4]. Pentney J granted the application pursuant to s 52, which provides that:

52 The Federal Court has jurisdiction, on the application of the Commissioner or of any person interested, to order that any entry in the records of the Patent Office relating to the title to a patent be varied or expunged.

While the matter was routine, one point caught my eye. Pentney J remarked that:

[5] In prior cases involving the addition or deletion of an inventor’s (or co-inventor’s) name, the Court has considered the test set out in sub-sections 31(3) and (4) of the Act, which relate to the addition or removal of applicants to a pending patent application: Plasti-Fab v Canada (Commissioner of Patents), 2010 FC 172, at para 14; Qualcomm Inc v Canada (Commissioner of Patents), 2016 FC 1092; Gilead Sciences Inc. v Canada (Commissioner of Patents), 2019 FC 70. The pertinent part for the purposes of this application requires that the Court be satisfied: (i) that the person should be joined as a co-inventor; and (ii) “that the omission of the further applicant or applicants had been by inadvertence or mistake and was not for the purpose of delay” (sub-section 31(4)).

Note the careful language. Pentney J did not say that the test for changing inventorship under s 52 is the same as under s 31(3)(4), but rather that the Court “has considered” that test. Nonetheless, on the facts he applied the s 31 test, and there seems to be a creeping tendency to import the s 31 test into s 52.

While the point didn't make any difference on the facts, it seems to me that it would be wrong in principle to hold that the s 31(4) test applies under s 52 as well. (And I would repeat that Pentney J did not so hold.) Section 31(4), relevant to adding an inventor, provides as follows:

(4) Where an application is filed by one or more applicants and it subsequently appears that one or more further applicants should have been joined, the further applicant or applicants may be joined on satisfying the Commissioner that he or they should be so joined, and that the omission of the further applicant or applicants had been by inadvertence or mistake and was not for the purpose of delay.

The second part of the test stated by Pentney J in [5], that “that the omission of the further applicant or applicants had been by inadvertence or mistake and was not for the purpose of delay,” is drawn directly from s 31(4), as Pentney J noted. Section 52 is much broader on its face. As Layden-Stevenson J noted in Micromass:

[15] The powers conferred on the Court under section 52 are very broad. In Clopay, Cameron J. described section 54 (now section 52) of the Act in the following manner:

...I think, therefore, that s. 54 was enacted so as to enable the rectification by the Court of the records in the Patent Office relating to title in order that the party or parties actually entitled to the grant or to be registered as to the assignees of the patent, might have their rights properly recorded (p. 235).

There is no suggestion in Micromass, or in s 52 itself, that s 52 is in any way restricted to cases of inadvertence or mistake.

The notion that the test from s 31 is relevant to s 52 appears to stem from Plasti-Fab 2010 FC 172 [14], where O’Keefe J, after quoting at length from Micromass, stated “Therefore this Court may, in place of the Commissioner, engage the test set out in subsection 31(4) to determine if an individual should be joined.” With respect, there is nothing whatsoever in the passage quoted by O’Keefe J, or anywhere else in Micromass, to suggest that s 52 engages the test set out in s 31(4).

Other cases have considered the relationship between the provisions. In Gilead 2019 FC 70 [2], Grammond J stated that “This Court has said, however, that the criteria found in section 31(4), which governs the addition of applicants to a patent application, are relevant to an application under section 52,” citing Micromass and Pasti-Fab. Plasti-Fab does stand for this proposition, via the questionable interpretation of Micromass, just discussed, but Micromass does not.

Qualcomm #1 2016 FC 499 (blogged here) concerned an application under s 52 to remove a co-inventor from a granted patent. Simpson J considered s 31(3), which relates to the removal of an named inventor from an application, where (i) “it . . . appears that one or more of them has had no part in the invention,” and (ii) the Commissioner is satisfied “by affidavit that the remaining applicant or applicants is or are the sole inventor or inventors.” Simpson J stated at [7] that “I am satisfied that the first part of the test in [s 31(3)] is relevant” under s 52; but this is only to say that a named inventor should not be removed unless they are in fact not an inventor. No doubt the Court would have come to that conclusion under s 52, even without consulting s 31(3). She then went on to say that the second part of s 31(3), the requirement for an affidavit, did not apply [8]-[11]. Thus, Qualcomm #1 is authority for the proposition that the test from s 31(3) is not applicable under s 52, though some of the same considerations may be relevant.

In Qualcomm #2 2016 FC 1092 (blogged here) Southcott J noted that Simpson J in Qualcomm #1 had held that affidavits were not required under s 2, but he accepted them nonetheless given that they had been sworn and introduced into evidence. He did not suggest that he disagreed with Simpson J’s holding.

It does seem entirely reasonable to say that the test from s 31 may be relevant under s 52, in that whenever the more restrictive requirements of s 31 are met, it will almost always be appropriate to order the correction under s 52. But on the face of the Act, and in light of Micromass and Qualcomm #1, it seems to me that the opposite is not true; s 52 might properly be invoked even in circumstances where the test from s 31 is not satisfied.

This is a longer post than is perhaps necessary given that the issue discussed made no difference on the facts. But I am concerned that by slow accretion of obiter statements it might become accepted that the s 31 test is required under s 52, thereby effectively reading a restriction from s 31 into s 52. I appreciate Pentney J’s careful language on this point, noting that prior decisions had considered the test from s 31, without saying that it was necessary under s 52.

Uncontested Change of Inventorship

Qualcomm Incorporated v. Canada (Commissioner of Patents) 2016 FC 1092 Southcott J
            2,630,594

This decision concerned an uncontested application by Qualcomm pursuant to s 52 of the Act to correct the name of the inventor on the 594 patent by adding the true inventor and deleting two incorrectly named inventors. In a previous Qualcomm decision, 2016 FC 499 (blogged here), Simpson J held that the affidavits that would be required to amend the inventorship for a pending application under s 31 are not strictly required under s 52. Qualcomm nonetheless provided affidavits, out of an abundance of caution [13], [14].

Qualcomm also sought to have certain documents recorded against the 594 patent, including a copy of replacements sheets for the PCT Declarations of Entitlement [1]. Southcott J declined to grant that relief, but without actually holding that the FC does not have the jurisdiction to grant such relief under s 52. Qualcomm did not urge the point very strongly, acknowledging that it would have other means have filing the replacement documents. Accordingly, in the absence of case law holding that s 52 authorized the FC to order the recording of documents, Southcott J declined to grant the relief, without any express holding as to whether he had the necessary authority to do so [16]-[17].

Affidavit from Remaining Inventors Not Required When Correcting Inventorship of Granted Patent

Qualcomm Incorporated v Canada (Commissioner of Patents) 2016 FC 499 Simpson J
            3,860,309

Qualcomm applied for an order pursuant to s 52 of the Act, to delete Mr Palanki as a co-inventor on the ‘309 patent. The application was not opposed by CIPO and Mr Palanki consented to the order sought on the basis that he was not a co-inventor. The only difficulty is that some decisions, in particular Imperial Oil 2015 FC 1218 (blogged here) and Segatoys 2013 FC 98, had suggested that in deciding whether to remove a co-inventor, the Court should follow the test set out for the Commissioner of Patents in s 31(3) [5], which provides that when a joint applicant for a patent is removed as not being an inventor, the prosecution may be carried on by the remaining applicants “on satisfying the Commissioner by affidavit that the remaining applicant or applicants is or are the sole inventor or inventors” [6]. In this case, even though it was uncontested that Mr Palanki was not an inventor, there was no affidavit that the remaining named inventors were the sole inventors. Simpson J held that the affidavit in question is essentially a housekeeping requirement to promote the efficient processing of pending patent applications in the Patent Office and it is not required when an issued patent is being considered by the Court under s 52 [9]-[10]. She therefore granted the application. I note that this is not inconsistent with Imperial Oil and Segatoys in the result, since in both of those cases the application was also granted: neither was a case in which an application to amend under s 52 was refused for failure to provide the affidavit in question.

Section 52 Allows Correction of Ownership

Gray Manufacturing Company, Inc. v. Canada (Attorney General) 2016 FC 55 Shore J
            2,406,340

This was an uncontested application under s 52 of the Act, which gives the Federal Court broad jurisdiction to vary or expunge Patent Office records related to the title to the patent, to correct the ownership of the patent. The inventor had intended to assign his rights in the invention to Gray Automotive Products Co., but the assignment document (which was attached to the patent application) inadvertently specified Gray Automotive Products, Inc., and the Patent Office consequently registered it against that name [4]. The mistake was unintentional, made in good faith, and, was made without any attempt to mislead or cause delay [10]. Shore J held that the broad jurisdiction conferred by s 52 includes the power to vary errors relating to the ownership of a patent [9] and he ordered the appropriate correction.

For a slightly more extended discussion of s 52, see here.

Update: Here is an interesting note by Ken Bousfield describing Gray Manufacturing as symptomatic of a secular shift in CIPO's attitude towards clerical errors.

Speculative Prejudice to Third Parties in Reinstating a Patent

Repligen Corporation v. Canada (Attorney General) 2012 FC 931 Near J

Repligen’s former patent agent transposed two digits of Repligen’s patent 1,341,486, to 1,314,486, when it engaged Computer Patent Annuities to pay maintenance fees on the patent. Consequently, the fees were paid against the “14" patent, which was held by Rolls-Royce. Sometimes the fees would be paid first on behalf of Rolls-Royce, sometimes on behalf of Repligen. The second payor would to be told that fees had already been paid and would request a refund. While duplicate fees were paid on the "14" patent, no fees were paid on Repligen’s “41" patent and CIPO sent a Maintenance Fee Notice to Repligen’s former agent advising that the patent would lapse unless payment was made within twelve months. Repligen’s former agent did not respond [5], and the “41" patent lapsed.

Repligen then switched agents. Its new agent attempted to have the patent reinstated, arguing that the Commission should exercise her discretion to correct the clerical error under s 8 of the Act and reinstate the patent. The Commissioner refused, primarily on the basis of the lengthy delay and the potential for prejudice to third parties who might have relied on the register showing the patent as lapsed [#1 59].

In Repligen #1 2010 FC 1288 Lemieux J held that the Commissioner had not properly exercised her discretion. A crucial point was that the Commissioner’s concern regarding prejudice to third parties was speculative as she had no evidence of third party prejudice. Lemieux J contrasted this with Bristol-Meyers (1998), 82 C.P.R. (3d) 192 (FCA) in which two other companies had filed priority claims for similar drugs [#1 60(f)]. Lemieux J noted the “catastrophic consequences” of the loss of patent rights [#1 59], and insisted that the Commissioner take into account the impact on Repligen in deciding whether the patent should be reinstated [#1 60(a)].

I am inclined to think that the Commissioner was right on this point. There may be many cases in which a third party would have relied on the lapse of the patent in a way that could not come to the attention of the Commissioner. For example a third party might have done patent clearance search before deciding to invest in developing a new product and then proceeded after having determined that its product would infringe the Repligen patent, but that the patent had lapsed. Given the nature of third party rights, it seems likely that it would be only in relatively unusual circumstances, such as those in Bristol-Meyers, that the Commissioner could become aware of the prejudice to third parties. It is true that the consequences to Repligen of the loss of its patent rights can be catastrophic, but it is also true that infringement of those rights could be catastrophic to a third party.

In any event, this was not the view of Lemieux J, who remitted the matter to be reconsidered by a different official in the Patent Office. The Commissioner once again refused to correct the error, and Repligen once again sought judicial review, leading to this decision in Repligen #2. This time the Commissioner focused on potential prejudice to Rolls-Royce. The Commissioner argued that “‘[c]orrecting Repligen's error today could retroactively have the effect of causing the Rolls-Royce patent to have expired in 2008' since the payments originally applied to it would automatically turn over to the Repligen Patent” [# 2 16]. Near J rejected this argument, rightly in my view. Fees were always paid in respect of Rolls-Royce’s patent, which never lapsed and has now expired [# 2 31]. I see this argument by the Commissioner as a way of attempting to point to an actual prejudice rather than a speculative prejudice in order to satisfy the direction by Lemieux J. It is an unsatisfactory attempt, as the Commissioner identified the real issue correctly in the first instance; the real question is whether the Commissioner should consider harm to third parties who may have relied on the lapse, unbeknownst to the Commissioner. Lemieux J, and now Near J, have held that it should not.

The matter had been remitted to the Patent Office once again. Even if the patent is reinstated, I am not sure Repligen’s troubles are over. Loss of patent rights is not “catastrophic” if those rights are never exercised. But if Repligen has the patent reinstated and then brings an infringement action against a competitor who can show that it did in fact rely on the lapsed patent, there is a a very strong argument that that defendant should have some kind of equitable defence. If the defendant can say “I searched for patents before undertaking my business, I found this patent that I would have infringed, I determined that it had lapsed, and I then invested $10 million in reliance on that,” it would be a serious failure of the patent system to allow Repligen to enforce against that defendant.