In Teva v. Pfizer Canada, 2017 FC 526, the Federal Court reaffirmed and reissued a judgment awarding Teva a section 8 damages award in excess of $125 million relating to the drug EFFEXOR XR® (venlafaxine). This decision offers insight into the legal limits on what inferences can be drawn about a generic’s ability to source sufficient drug supply in the but-for damages world.
The key issue at this redetermination was whether Teva would have had and could have had access to sufficient quantities of venlafaxine at the relevant time to support its notional sales in the but-for world that gave rise to its significant section 8 damages award.
At trial, no current or former employees testified from Teva’s venlafaxine supplier, Alembic. Teva did not adduce any direct evidence from Alembic. Instead, Teva relied upon the testimony of its witness Mr. Major, who was a former executive at Teva. Following the redetermination after reversal on the basis of improper reliance on hearsay, the Federal Court restored its initial judgment on the basis of “reasonable and logical inferences” from the following key evidence that Teva did present at trial:
1) As to the “would have” element, the Federal Court relied on a License and Supply Agreement between Alembic and Teva where Alembic committed to “supply exclusively to [Teva] … all of [Teva’s] requirements of Product.” Wyeth (subsequently Pfizer) acknowledged that this showed the intention of the parties at the time of the agreement. Furthermore, Alembic in fact manufactured venlafaxine prior to January 2006 for Teva’s regulatory submission. Justice Zinn held that “there can be no better evidence” of Alembic’s willingness to supply Teva with venlafaxine in the but-for world than its supply of venlafaxine in the real world”.
From the reasons it does not appear that the “would have” argument was seriously contested. Rather, the focus of the reasons was on whether there was admissible evidence that Alembic “could have” supplied venlafaxine, and if so whether it could do so at the amount required to satisfy the section 8 damages award.
2) The “could have” element required Teva to prove that it could have obtained sufficient venlafaxine to make nearly 248 million capsules sales from January 1, 2006 to August 1, 2007. The trial judge inferred that Alembic’s production capacity at the relevant time was one billion capsules annually based in part upon Mr. Major’s evidence. At trial, after being shown an e-mail from Alembic describing its capacity, Mr. Major testified that: “A billion is reasonable based on what I recall from having visited the plant…” Notably, the Federal Court of Appeal found the trial judge to have erred at first instance in relying on that same e-mail as proof of Alembic’s capacity. However in the redetermination, the trial judge reasoned that even though the e-mail itself was unauthenticated and inadmissible, it could serve as “aide memoire, or trigger” for Mr. Major given his first-hand observation of what he saw during his inspection in 2004.
The trial judge further accepted that there was no evidence as to how much available capacity Alembic had at the relevant time. However, the trial judge inferred that Alembic would have had sufficient available capacity based on its subsequent agreement with Teva to supply the larger USA market, and Alembic’s contractual obligation to supply Teva’s needs for the Canadian market.
In drawing this inference, the trial judge rejected Wyeth’s argument that Teva failed to prove that Alembic did not have other contractual commitments for its capsule machinery that would have prevented it from meeting Teva’s needs. Justice Zinn held that in the face of the supply agreement between Teva and Alembic, the evidentiary burden shifted to Wyeth to lead evidence that the contract would not or could not have been fulfilled according to its terms – for which Wyeth had none.
This decision serves as a cautionary tale on the application of the rules of evidence to section 8 cases and the risks of relying on second hand information and inferences where first-hand accounts are not adduced at trial. The extent to which first persons may have the burden to disprove inferences of capacity will be closely scrutinized in future decisions.