When success does not mean survival – Generics knock out Sanofi’s cancer drug patent at the UPC

2025 Week 51

Generic pharmaceutical companies scored their first big win at the UPC last week, with the Munich Local Division (LD) revoking Sanofi’s patent EP 2493466 B1 for a medical use of the taxane drug cabazitaxel.

In combination with prednisone or prednisolone, cabazitaxel is used for treating metastatic castration-resistant prostate cancer (mCRPC) in patients previously treated with another taxane, docetaxel. Sanofi’s product Jevtana® has faced competition from generics since its regulatory exclusivity expired in March 2021.

Sanofi had tried to protect its market position by enforcing this patent in the UPC, against a number of generics companies, who counter-claimed for revocation of the patent.

In revoking the patent for lack of inventive step, the Munich LD diverged from the Board of Appeal of the European Patent Office, where the patent had recently survived a twelve-opponent attack (decision T136/24). The Munich LD chose instead to follow the earlier decision from the Judicial Court of Paris, who in September 2024 had found the French part of the EP patent to be invalid, in light of prior art disclosures relating to the ongoing Phase III ‘TROPIC’ clinical trial.

Different frameworks, different outcomes?

UPC watchers cannot have missed the fact that the UPC Court of Appeal (CoA) recently set out their definitive test for the assessment of inventive step (UPC Weekly 2025 Week 48).

Surprisingly to some, it was not the EPO’s long-established ‘problem-solution approach’ (a structured approach which starts from a single ‘closest prior art’ disclosure). Rather, the CoA endorsed a more holistic, ‘next step’ approach. This approach is now consistently being applied by lower UPC courts (see UPC Weekly 2025 Week 49).

The party line from both the UPC and the EPO, since the CoA guidance was issued, has been that the two approaches are actually not very different and, importantly, that when applied properly they should generally lead to the same conclusion.

But in this case, they did not. It is worth asking: what factors led to this divergence of opinion, and would the outcome in the UPC case been different if the court had approached it using the EPO’s problem-solution framework?

TROPIC: Trial or Tribulation 

The Munich LD and the EPO Board of Appeal both considered the most realistic starting point in the prior art to be disclosure relating to the ongoing Phase III clinical trial, investigating the efficacy of cabazitaxel plus prednisone in the relevant patient group. This was referred to as the ‘TROPIC’ trial, the results of which were disclosed for the first time in the patent application. The trial protocol described the indication to be treated, patient population, treatment regimen, and primary endpoint (overall survival), and was publicly available before the patent’s filing date.

Both the UPC and EPO decisions made it clear (following established EPO Board of Appeal case law) that the mere existence of a phase III clinical trial should not automatically create an expectation of success in respect of the treatment under investigation: this must be evaluated on a case-by-case basis, from the standpoint of the skilled person. The Munch LD explicitly confirmed that:

"[t]he fact that a clinical trial is ongoing does not in itself create a presumption of success. The expectation of success must be assessed on the basis of all the circumstances of the case, including positive and negative indicators”
(p. 33 of the decision),

with the EPO Board of Appeal similarly stating that:

“[t]he mere fact that a phase III clinical trial is in progress does not automatically mean that the skilled person would have a reasonable expectation of success. This has to be evaluated in light of the available data and the specific context, weighing both positive and negative factors”
(T136/24, reasons 7.13).

Where they differed, however, was in what ‘success’ would have meant to the skilled person in this case and also in their weighing up of the positive and negative pointers towards that success in the prior art.

Measures of Success

The EPO Board of Appeal, starting from the clinical trial disclosure as “closest prior art”, assessed whether the skilled person would have expected the trial to succeed (T136/24, reasons 7.13).

They formulated the objective technical problem in view of the closest prior art as “to put into practice the effective treatment of prostate cancer with cabazitaxel in co-administration with prednisone” in the relevant mCRPC patient population.

In considering whether the claimed solution was obvious, the decisive factor was “whether the person skilled in the art would have had a reasonable expectation of success with regard to the… TROPIC study” (T136/24, reasons 7.10).

For the Board, ‘success’ in this context meant the achievement of the trial’s primary endpoint, namely an increase in overall survival.

Based on the underlying problem identified in the patent the Munich LD formulated the objective technical problem more broadly, as the provision of “a therapeutic option” for the relevant patients, criticising the EPO Board of Appeal’s formulation of the objective technical problem as containing parts of the solution and therefore involving hindsight.

Crucially, the Munich LD considered that the question of success rested not only on whether the trial would achieve its primary endpoint, but whether the skilled person would expect the approach described in the trial to provide any therapeutic benefit at all, including mere palliative treatment as well as survival improvement (decision, pp. 33–34).

The Munich Local Division justified its broader formulation by emphasising that the patent itself disclosed multiple endpoints (PSA level, tumour progression, pain relief, progression-free survival) as accepted indicators of therapeutic effect, not survival alone. The court stated:

“It is therefore irrelevant whether or not the patent is limited to survival alone … these different indicators are indissociable and it is not possible artificially to isolate different aspects of the therapeutic effect.”

Balancing Act

Both the Munich Local Division and the EPO Board of Appeal agreed that expectation of success must be evaluated case by case, considering all relevant factors. However, they gave very different weight to the same indicators.

The EPO Board of Appeal, in their decision, more strongly emphasised the negative pointers.

They noted the absence of any Phase II clinical data in prostate cancer, the high toxicity profile of cabazitaxel, and the failure of other taxanes in similar settings. These technical factors, in their view, outweighed positive signals such as the existence and continuation of the Phase III trial, and the preclinical data which they considered to be too limited to allow any predictions of clinical success. The Board concluded that the skilled person would have had “at most a hope” of success, not a reasonable expectation (T136/24, Reasons 7.13).

The Munich LD, by contrast, were more persuaded by the positive indicators.

They considered the advanced stage of the TROPIC trial—running for three years without interruption—as a strong sign that the approach was promising. In particular, the court inferred that the sponsor’s decision to continue the trial supported expectation of success, taking into account the significant cost of such trials. They placed decisive weight on the fact that the TROPIC trial, at the priority date, had reached near-completion without interruption, interpreting this as a strong pointer towards success, regardless of whether success meant survival or broader palliative benefit:

“What can … be derived from the fact that the Tropic trial had been approved and had been in progress for three years … is that at least the sponsor … had not considered it disappointing … This supports the expectation of success … for the use of the combination of cabazitaxel and prednisone.”

This contrasts with the EPO Board of Appeal, who did not lend much weight to the decision of the sponsor to continue with the trial, characterising this as a “business decision” based on “underlying financial risk/benefit analysis” from which the skilled person would not derive any concrete technical information about whether the trial was likely to succeed.

The Munich LD also gave significant weight to prior art linking cabazitaxel’s performance in breast cancer to potential efficacy in prostate cancer, describing this connection as “clear” (pages 34, 43–44 of the decision). Negative factors such as toxicity were acknowledged, but in their view did not outweigh these positive signals.

In coming to their conclusions on the positive and negative pointers, the Munich LD relied not only on the disclosure of the relevant documents but also on the testimony of the parties’ experts. Both experts had been specifically questioned (together, in a so-called ‘expert hot tubbing’ hearing) about whether there would have been a reasonable expectation of a positive outcome from the TROPIC trial. 

Would the UPC outcome have changed with a different approach?

Based on the above, it seems unlikely that the Munich LD would have reached a different conclusion if they had followed more closely the EPO’s approach and formulated a narrower technical problem (based on the ‘closest prior art’ and hence, potentially, focussed more narrowly on the stated endpoints of the TROPIC clinical trial).

Even if the Munich LD had adopted a narrower definition of ‘success’ (requiring improvement in overall survival, as in the EPO case) the reasoning in their decision indicates that the court already considered survival improvement to be part of the reasonably expected outcome of the TROPIC trial. They concluded that:

“The encouraging results of cabazitaxel on prostate cancer and on resistance to docetaxel could give [the skilled person] expectation of a favourable effect on survival, even though these results related only to indicators other than survival alone” (decision, page 40).

The Munich LD’s analysis of expectation of success was anchored in the trial’s advanced stage and absence of negative signals, rather than in the breadth of the endpoints considered.

It noted that survival was part of the “inseparable whole” of the expected therapeutic effect and that the skilled person could reasonably expect a favourable effect on survival.

Therefore, adopting a slightly narrower problem to be solved based on the identified ‘closest prior art’ would not have removed the court’s main positive pointers (trial continuation, prior data, same taxane class). Their conclusion of obviousness would therefore likely have remained the same.

Conclusions

Arguably, the different frameworks used by the EPO and the UPC led to different objective technical problems being formulated, with the EPO’s problem-solution approach resulting in a narrower formulation, closely tied to the endpoints stated in the trial protocol, whereas the UPC approach looked more broadly at the technical problem as set out in the patent. But was that the decisive factor?

Probably not. Although the Munich LD criticised the Board’s formulation as involving hindsight, the use of hindsight would usually make an invention seem more obvious, not less, so this should not explain the divergence in this case.

Given their views on the various ‘pointers’ in the art, it seems likely that the UPC judges would have come to the same conclusion on obviousness, even if they had considered improvements in overall survival as a necessary element of success, because it was clear that they thought the skilled person would expect the PhIII trial to succeed.

Ultimately it came down to the balance of the pointers (positive v negative) for expectation of success and how these were viewed. For the UPC judges, the balance tipped one way, for the EPO Board of Appeal members, the other.

With both tribunals being clear that each case must be assessed on its facts, the door remains open for medical use patents based on clinical trials to be upheld by the UPC, but patentees in the pharma space will likely be more cautious than ever in view of this decision. It remains to be seen, of course, whether the UPC Court of Appeal will agree with the way their test was applied by the lower court.