The article was first published on the Kluwer Trademark Blog

It is understandable why (most) trademarks owners do not like the five-year dependency period of an international registration since in case of “central attack” or refusal of the basic application, they face cancellation of their international registration. However, it’s less understandable why the judges of the EU General Court (GC) do not like it, unless perhaps, the sad truth is that they do not quite grasp the concept.

In the quite remarkable decision on Case T-459/22 issued on 3 May 2023, the GC seems to have mixed up the concept of cancellation following the abandonment of the US basic trademark application during the five-year dependency period, with a non-renewal.

The facts are straightforward.  Biolark Inc. designated the European Union through the international trademark no. 1453505 “BIOLARK” based on a pending US trademark application.

Laboratorios Ern. filed an opposition. The EUIPO rejected the opposition, the Board of Appeal confirmed this on 22 April 2022, and Laboratorios Ern. filed an action to the GC on 21 July 2022. In the meantime, on 14 March 2022, the US trademark application for BIOLARK was considered abandoned by the USPTO because no Statement of Use had been filed by Biolark Inc.  As a consequence, the USPTO informed WIPO which duly recorded the cancellation of the IR on 8 November 2022.

According to the GC decision, EUIPO “informed the Court that the international registration which was the subject of the application for the grant of the protection of the EU trade mark system, relating to the mark applied for had not been renewed. Consequently, according to EUIPO, the present action has become devoid of purpose” (cf. at §14).  It is hard to believe that EUIPO made such a fundamental mistake, but we must trust the GC to report correctly what the parties’ briefs say.

Surprisingly, the GC did not look into this further and did not realize that cancellation of the international registration was due to the abandonment of its US basic application. The GC therefore actually wrote in its decision that “It is, moreover, common ground that the international registration was not renewed as from 24 November 2022”. Based on this wrong understanding, it rejected EUIPO’s contention that the action had become moot, and stated: “given that the application for registration was filed with effect from 6 January 2019, that the non-renewal of the mark applied for did not take place until 24 November 2022 and that there is nothing to suggest that that non-renewal would have ex tunc effects, the contested decision, if it becomes final, would have the effect of rejecting the applicant’s opposition and thus removing the obstacle to registration of the mark applied for for the period claimed, until its expiry. The mark applied for would thus be regarded as eligible for registration between the date claimed, namely 6 January 2019, and the date of its expiry, [emphasis added] namely 24 November 2022”.

This is not the right place to enter into detailed discussion on whether the dependency clause (art. 6.3 of the Madrid Protocol) produces ex tunc or en nunc effects. However, under EU laws a trademark application is not able to confer exclusive rights per se, unless it reaches registration.  So one wonders which legal effects could be tied to the question whether an EUTM application was lost sooner or later in the application process (and the same applies to an application for grant of protection of an IR).

On the other hand, US trademark Law (Lanham Act) provides that in order to a confer to a trademark exclusive right the trademark must be used in commerce, with the consequence that if the applicant of an intent-to-use trademark application does not furnish evidence of use within the prescribed period, the trademark application is considered to be abandoned, and no exclusive rights are granted. Thus, in the case at issue it would be surely weird that an IR based on a pending US application which was never granted may have (temporal) effects greater than the base application ever had.

But even if the dependency clause produces effects ex nunc, and not ex tunc, that does not excuse the GC to mistake central attack consequences with non-renewal. So, does anyone need a quick 101 crash course on the Madrid Protocol?