Thalidomide Is Not an "Innovative Drug"

Celgene Inc. v. Canada (Health) / THALOMID 2013 FCA 43 Gauthier J, Sharlow J concurring: Nadon JA dissenting, rev’g 2012 FC 154 de Montigny J (blogged here)

Text, context and purpose are all important in statutory interpretation, and in hard cases, where these considerations do not all align, the result may depend on which element is emphasized. TheTHALOMID litigation over the interpretation of the term “innovative drug” under the Data Protection Regulations, shows three different judges weighing these factors differently. In the FC, de Montigny J focused almost entirely on what he saw as the purpose of the Regulations. In the FCA, Gauthier J, for the majority, adopted an approach centered primarily on the text, while Nadon J, dissenting, had a relatively balanced approach, and relied on text, context and purpose. The result on the facts is that the FCA reversed de Montigny J decision that thalidomide should be listed. The more general consequence is a negative one. de Montigny J’s interpretation would have had very wide ramifications for the interpretation of “innovative drug.” The FCA approach turns more on the particular and very unusual facts of this case. This is true in respect of both the reasons of Gauthier J for the majority, and Nadon J in dissent.

The question in this case is whether thalidomide is an “innovative drug” under the Data Protection Regulations, and so eligible for listing on the Register of Innovative Drugs. Thalidomide was originally launched 1957, and two companies received regulatory approval to sell thalidomide in Canada in 1960 and 1961. In 1962, after its teratogenicity became known, the Department of Health ordered thalidomide's permanent withdrawal from the Canadian market. The withdrawal letters sent to the manufacturers stated in part that “[w]ith the withdrawal of this acceptance, thalidomide returns to the status of a new drug” [6].

Over the past 20 years Celgene has developed various applications of thalidomide. In May of 2009 Celgene filed an NDS seeking approval for thalidomide, which it markets as THALOMID, for use in the treatment of multiple myeloma. The NOC was issued in August of 2010 [12]. However, the Minister of Health advised Celgene that THALOMID was not eligible for data protection because an “innovative drug” is defined in C.08.004.1(1) of the Data Protection Regulations to mean “a drug that contains a medicinal ingredient not previously approved in a drug by the Minister,” and thalidomide had been previously approved by the Minister in the 1960 and 1961 approvals. Celgene sought judicial review of this decision.

In the FC, de Montigny J held that thalidomide is eligible for listing. As discussed in my post on that decision, rather than relying on the fact that the previous approval for thalidomide had been withdrawn, de Montigny J’s interpretation emphasized that “the purpose of the [Data Protection Regulations] is clearly to encourage and reward innovation by protecting the data an innovator must generate to obtain approval for a drug” [36], and in this instance, the new approval of thalidomide had relied on very extensive new data.

On appeal, while Nadon J, dissenting, noted that a purposive interpretation is helpful, both he and Gauthier J for the majority, focused on the text of the Regulations. Gauthier J stated that a part of de Montigny J’s purposive analysis which suggested that a new use of an approved ingredient should qualify as an innovative drug, was not really argued before him, and his comments on this point “should therefore be given no precedential value” [38]. Thus the broad implications of de Montigny J’s decision that I remarked on in my previous post, have been erased.

As Gauthier J put it, “the true question before us [is,] Is thalidomide a ‘medicinal ingredient not previously approved in a drug by the Minister’?” [40] Gauthier J’s reasoning was quite straightforward. Because of the withdrawal of approval, thalidomide was considered a “new” drug, but “new” and “innovative” are distinct concepts under the Regulations. She pointed out that Celgene “did not argue that this drug was never approved (i.e., that the approval was null ab initio)” [41]. Therefore, on the face of it, thalidomide was “previously approved.” In her view, to hold the contrary would “require [the court] to construe ‘previously’ as meaning ‘currently’ or to read in the words ‘and currently’ before the word ‘approved’ in the definition” [44].

The remainder of Gauthier J’s reasoning was primarily devoted to showing that considerations of purpose could not overcome this textual hurdle. In particular, the Regulations are intended to implement TRIPS and NAFTA, and these do not contemplate protection for all data “ the origination of which involves a considerable effort” but only for such data that is also related to a “new” entity. Gauthier J was also concerned about the line drawing implications of a broader interpretation: “How many other changes in the regulatory requirements throughout the years since then could be argued to be significant enough to warrant another exception to the rule?” [62]

At a technical level, the key difference between Gauthier J and Nadon J is Nadon J held that “[f]or all intents and purposes, the manner in which thalidomide has been treated has amounted to a nullification of any previous approval” [72] (see also [73], [90]), and “[t]his conclusion therefore allows data protection to be extended to thalidomide: the “previously approved” condition in the definition of innovative drug has not been met” [90]. This contrasts with Gauthier J’s statement that Celgene “did not argue that this drug was never approved (i.e., that the approval was null ab initio)” [41]. This was the only place in which she addressed the nullification argument, and it not entirely clear whether Gauthier J was of the view that the approval was not nullified in law, or whether she was simply unwilling to consider the argument as it had not been made by Celgene. 

Nadon J also gave more weight to contextual and purposive considerations. For example, he noted that the general scheme of the PM(NOC) regulations is that drugs are either innovative or generic, and as between the two, thalidomide falls more appropriately into the former category. He also argued that “previously approval” could not be interpreted literally, since the FCA in ELOXATIN 2012 FCA 106 (blogged here) held that approval under the SAP program does not constitute previous approval for the purposes of the Data Protection Regulations. Nadon J was also unpersuaded by the line drawing argument. He emphasized the completely unique history of thalidomide, and concluded that “[a]ny precedent set by this decision is necessarily narrow in scope and does not generate these slippery slope concerns” [101].

I think it is fair to say that in considering text, context and purpose, Nadon J’s approach is more balanced than either de Montigny J, who considered purpose almost exclusively, or Gauthier J, who primarily focused on the text. This is not to say that his approach is best for that reason. As McLachlin J stated in Canada Trustco Mortgage Co. v. Canada, 2005 SCC 54 [10] “When the words of a [decision] are precise and unequivocal, the ordinary meaning of the words play a dominant role in the interpretive process.” But then the question is whether the words “previously approved” are precise and unequivocal. Gauthier J evidently thought they were. Perhaps the only general lesson here is that truly difficult questions of statutory interpretation are not as uncommon as we might hope.