The EU General Court (Eighth Chamber) has confirmed that an applicant who wants a new distinguishing characteristic weighed during the technical examination of a plant variety has to do considerably more than write it on the form. In Adema Srl v Community Plant Variety Office, Case T-770/25, judgment delivered on 14 October 2026, the Court upheld the CPVO Board of Appeal’s refusal to take account of a characteristic the applicant had flagged in its technical questionnaire but never supported with data. According to IPKat’s report of the judgment, the Court held that a request of that kind in the application neither obliges the Office to include the characteristic in its examination of the distinctness, uniformity and stability criteria, nor predetermines what that examination will find.
The underlying facts are a fifteen-year cautionary tale. The application concerned an apple variety named ‘Roat’, assessed against the reference variety ‘Jeromine’. A first technical examination ran from 2011 to 2016 and produced a finding of no distinctness; the Board of Appeal agreed in 2016 that the assessment was flawed because the two varieties had been grown on different rootstock, and the examination was repeated with new plant material. That second examination, delegated as usual to a national authority — here the French seed and variety testing body GEVES — again reported that the two varieties could not be clearly distinguished, with a final report in 2023. Testing followed the CPVO’s apple protocol TP/14/2 for Malus domestica, derived from the UPOV guidelines, in which the nearest listed trait is Characteristic 23, “Young fruit: extent of anthocyanin overcolour”. The applicant had asked for something subtly different: the early development of the overcolour, meaning the timing of the fruit’s reddening rather than its extent. The Board of Appeal’s answer was that the applicant needed to show this was fundamentally different from Characteristic 23, and it had never identified the point in time at which the two varieties’ colouring actually diverged. The General Court accepted the point of law that the CPVO President can add characteristics to test guidelines even after a technical examination has finished, but affirmed that this one had never been qualified as a new characteristic at all: the protocol requires a claim for a new characteristic to be backed by reliable technical data before a special test can be devised, and — as the Court noted — no such data was filed at any stage, either during examination or before the Board.
What this changes for breeders and their advisers
The technical questionnaire is not a reservation of rights. The most expensive assumption in this case was that mentioning a characteristic in the application preserved the ability to rely on it later. It does not. Distinctness is decided on the characteristics the protocol lists unless and until the applicant converts a mention into a substantiated request. If your variety’s commercial point of difference is a trait that falls outside the relevant TP protocol — a ripening or colouring timing, a storage behaviour, a resistance expressed under conditions the protocol does not create — then the questionnaire entry is the beginning of the work, not the end of it.
Supply the data before the examination is designed, not after it fails. The protocol route runs in one direction: reliable technical data first, then an assessment of whether a technically acceptable test procedure can be devised, then the special test. Data produced after a negative final report arrives cannot retro-fit a test that was never run. In practice that means the trial data, measurement methodology and observation timings that distinguish your candidate from the likely reference variety should exist in presentable form at filing. Where the Office tells you a special procedure is available — the CPVO said exactly that to this applicant in 2013 — treat that as a deadline with no reminder attached, because nothing in the system will chase it.
Interim reports are the window, and silence closes it. The Court’s reasoning placed weight on what the applicant did not do: it did not renew its request when GEVES’s interim reports showed the varieties were not separating, and it did not raise it during site visits. Every interim report on a DUS examination should therefore be read as a docketed decision point — either the standard characteristics are distinguishing your variety, or you re-assert the additional characteristic in writing, with data, on that report. Growing-trial calendars make this unforgiving: a missed cycle is a year, and in this matter two examinations consumed more than a decade.
Pick the reference variety fight early. Distinctness turns on comparison with varieties of common knowledge, and the reference variety is selected from the description the applicant itself supplies. Two of the three substantive outcomes here were driven by that comparison — the first examination was redone only because the rootstock differed between candidate and reference. Mapping the field of comparable varieties before filing, and understanding where a candidate genuinely separates from each of them, is the same discipline as any other pre-filing clearance exercise; the same portfolio analysis logic that decides which traits to claim and where applies to varieties as much as to patent families.
Do not let the PVR layer stand alone. IPKat places the decision alongside the EU’s recent regulation on new genomic techniques and the open question of how the innovative features of NGT plants will be recognised in a PVR system built around protocol-listed characteristics. Breeders working on traits that the protocols do not describe are therefore exposed twice over: the trait may be commercially decisive and still not count toward distinctness. That argues for a layered strategy in which patents, PVRs and contractual controls are planned together, and for checking the patent landscape around a trait and its enabling technology — an ordinary freedom-to-operate question — rather than treating the variety right as the whole of the protection.
One limit on how far to read the case. As IPKat notes, the General Court reached the opposite practical result in the SK20 decision, Case T-556/22, concerning an onion variety that does not make people cry: there the “no tear” characteristic did not need examining because the variety was already distinguishable on the standard protocol characteristics, and it was simply excluded from the description. The two decisions fit together as a single rule. If the listed characteristics already separate your variety, a novel trait is a marketing asset and not an examination issue. If they do not, the novel trait is the whole case — and under T-770/25 it only enters the case with technical data behind it.