"In 2004, the Court of Appeal for England and Wales held in Coflexip v Stolt that once a defendant had been found liable for infringement, it would still be liable to pay damages (when eventually assessed) even if the patent was held invalid in later proceedings with a (necessarily) different defendant. Subsequent cases, most recently the Virgin Atlantic v Premium Aircraft case, have reviewed this decision which remains part of English law. The cases have recently been reviewed in an article in MIP (MIP, Dec/Jan 2011 p20, Evans et al).Readers' thoughts and comments are keenly awaited.
This author's view is that the approach in Coflexip is undoubtedly the correct one. While the proposition may seem surprising, it is in fact entirely consistent with fairness in patent system. There are a variety of approaches that lead to this conclusion.
"In particular, I find symmetry a convenient tool by which to assess fairness. In Jeremy's scenario, there is no symmetric treatment of the parties - the first defendant would obtain multiple bites at the "damages cherry" until the patent is invalidated, whereas the patentee has no second chance once invalidity has been found. Turn the Tribunal B and A decisions around, and it is clear that as the law stands, tribunal A cannot hold the patent valid after tribunal B has held it invalid because once that has happened it is curtains for the patentee. So if the patentee is going to be fixed with finality, it is only fair that the same should apply to defendants.
In fact, of course, a defendant does benefit under the present system because while the damages liability remains - dealing with a crystallised liability in the past - the injunction into the future is nugatory because there is no longer a patent capable of being infringed. So the defendant already has had a leg up and doesn't need more.
It might pay the defendant in Tribunal B to drop the counterclaim for revocation and just pursue the invalidity defence. No inconsistent decision can then be created, and the B defendant is not placed in the invidious position of getting the A defendant off the hook by doing a better legal job. On the other hand, the scenario Jeremy preferred above is one that rewards the first defendant for the risk and expense borne by the second defendant and provides the patentee with no finality in litigation. It provides the first defendant with a second bite at the cherry that is not available to the patentee. It incentivises the first defendant to delay settlement or refuse to settle.
The key point in Coflexip is that the decision on damages is already a done deal. The only remaining question is quantum and it is just a question of timing and convenience that quantification of the liability is delayed to a later date. In other words, the damages inquiry is properly seen as teasing out the detail of an accrued right, and not deciding on new rights.
Which, in short, means Coflexip was rightly decided. Patent validity as a relative concept is already deeply embedded in the law - it is why there is no repayment of licence fees, settlement payments or accrued damages following a finding of invalidity. The risk of a patent being valid or invalid is a variable that is factored into all dealings with patents, and rightly so. It creates a value spectrum leveraging the perceived likelihood of validity, from rights that look decidedly dodgy and are valued and treated cheaply, to rights that look as good as you get and are valued highly. If payments were subject to retrospective correction after an actual finding of invalidity, then that raises risk and cost for the patentee which will tend to raise the immediate price for licensees for dealing with the patent, and thus make agreement more difficult to achieve and litigation more likely. It would be no more constructive to argue that historic payments should be retrospectively raised following a finding of validity, because the previous probabilistic assumptions about validity are resolved. The system is far better to stick with the treatment as is".