Sunday, August 20, 2006

Latest round of the heartburn wars: not about Nora Ephron

AstraZeneca LP v. TAP Pharmaceutical Products, Inc., 2006 WL 2338144 (D. Del.)

AstraZeneca sought a declaratory judgment that its “Better is Better” campaign for Nexium, which has now been discontinued (see remnant here), was not false or misleading under the Lanham Act. TAP, producer of Prevacid, counterclaimed and demanded a jury trial. (The court found that TAP wasn’t entitled to a jury trial because its damages evidence had been excluded and its only remaining available relief was an injunction.)

Nexium and Prevacid treat acid reflux with a similar method of action. The “Better is Better” campaign ran for half a year, late 2004 to early 2005, and included two TV commercials, print ads, website materials, and a pamphlet. The basic claim: “recent medical studies ... prove Nexium heals moderate to severe acid related damage in the esophagus better than the other leading prescription medicine.”

TAP didn’t contest the reliability of the two studies on which AZ relied, which showed a difference of 12.7% and 4.9% in healing damage at eight weeks compared to Prevacid (86.6% versus 73.9% and 82.4% versus 77.5%), with greater improvements the more severe the damage was. Rather, TAP argued that the ads conveyed several false and misleading messages: (1) Nexium is better for symptom relief, (2) Nexium is clinically better, when it’s only marginally better at healing, and the statistically significant difference is clinically meaningless, and (3) Nexium is better for everyone, when only a small minority of patients have “moderate to severe” damage.

The court rejected most of the parties’ challenges to each other’s experts, holding that they went to the weight of the evidence rather than its admissibility, for example TAP’s testimony about the difference between statistical and clinical significance. TAP offered survey evidence, which AZ sought to discredit. Dr. Thomas Dupont did two surveys, one based on a TV commercial and the other on internet advertising.

Claims (1) and (3) (better for symptoms/better for everyone): Based on combined responses to general (paraphrase: What did the ad say?) and specific (Did the ad say Nexium is better? What is it better at doing?) questions, Dupont concluded that 71.1% of respondents took away a message that Nexium provides superior symptom relief, while 32.1% understood that Nexium was better at healing damage. A final closed-ended question allowed respondents to choose whether the ad said Nexium was better than other medications for symptom relief, healing damage, both, or neither (and a don’t know option). In response to that question, 20.8% of respondents said Nexium was only better at healing damage to the esophagus, 15.8% said it was better only at treating symptoms of acid reflux disease, and 62.5% said it was better for both. Dupont, however, stated in his report that this last question was only included as a precaution, in case the answers to the open-ended questions were ambiguous. He felt that question six may have been either suggestive, overly complex, or both, and that the question was unnecessary because the answers to the open-ended questions were clear.

The internet survey included a question assessing materiality and its closed-ended question dealt with whether Nexium was better just for people with moderate to severe damage or for some other group. Before the close-ended question was asked, 11.2% of respondents who viewed the test ad said something about moderate to severe damage, while no respondents who viewed the control ad made such a statement. Dupont stated that these results "don't help us a lot in understanding whether or not consumers understood the limitation on the claim being made in the ad." In other words, it’s possible that more people understood the limitation on the claim than just 11%. In response to the closed-ended question, 38.5% of test cell respondents said that only people with moderate to severe damage to the esophagus were healed better by Nexium. However, 38.3% of control cell respondents, which did not contain the words moderate to severe, gave the same response. TAP’s rebuttal expert therefore opined that these questions should be excluded from the analysis. She believed that many respondents were merely guessing.

Perhaps unsurprisingly, similar percentages of test and control subjects indicated that the ad they saw would influence their purchase decision; very few gave the specific reason that Nexium heals damage better. But, of course, the real action is in the difference between the open- and closed-ended questions. The court agreed that the closed-ended questions prompted meaningless responses, and without the closed-ended questions it was impossible to interpret the answers to the open-ended ones. At most, all you can say is that at least 11% understood the limitation to moderate/severe damage, and somewhere between 0% and 89% were misled. Thus, the internet survey was excluded because it lacked probative value.

Independent of all this, AZ argued that a consumer could properly conclude that Nexium is superior overall because it’s superior in healing moderate to severe esophagal damage. TAP responded that there’s an established rule that a claim that’s only true with respect to a subset of patients is false when presented as an unqualified claim. But the court found that no such rule exists. My reaction: there are cases going both ways. The most logical way to divide them is to say that, if the subset is a small fraction of the total consumer population (people whose fatigue comes from iron deficiency rather than other factors, to take an example of a false claim once made for Geritol) or if the consumer can readily whether she’s a member of the subset (people who have migraines rather than normal headaches), unqualified claims that relate only to the subset are false. If the consumer could easily be a member of the subset and isn’t likely to find out, it can be worth trying the product and thus the unqualified claim isn’t explicitly false. If most people with heartburn don’t know the extent of their esophagal damage, then the court’s conclusion fits with prior cases. Given that Nexium is still useful for people outside the subset – unlike Geritol for people whose fatigue is from thyroid deficiency, say – it’s reasonable for consumers to conclude that, all other things being equal, Nexium is better because it’s better for moderate to severe damage.

Claim (2) (statistically significant versus clinically significant): TAP argued that Nexium’s statistical advantage over Prevacid was not clinically or therapeutically significant, and thus claiming an advantage was literally false. The court rejected this argument. The cases on which TAP relied involved situations where there was affirmative evidence that the difference touted by the defendant had no clinical significance. In Novartis Consumer Health, Inc. v. Johnson & Johnson-Merck Consumer Pharma Co., 129 F.Supp.2d 351 (D.N.J.2000), the defendant’s product was stronger in vitro than other products on the market, but in vivo results consistently indicated that greater strength didn’t improve efficacy in the body. Likewise, in Castrol, Inc. v. Quaker State Corp., No. 91 Civ. 8517(CSH), 1992 WL 47981 (S.D.N.Y. Mar.2, 1992), the defendant proved its product was faster flowing than other products, but tests of engine wear (which was the claim at issue) showed that faster flow, if anything, made things worse. Here, the studies on damage healing were relevant to consumer use.

One can imagine a case in which a statistically significant result was clinically irrelevant, though it might be hard to fit into current Lanham Act jurisprudence. For example, a rigorous enough study could decisively prove that Medication A is effective in 1% more patients than Medication B, or works on average one hour faster. Especially if Medication A is more expensive or has greater side effects, I find it hard to imagine doctors thinking that Medication A is clinically better. Now, expense and side effects are technically unrelated to efficacy and so ordinary Lanham Act law wouldn’t consider them in evaluating the truth of efficacy claims, but if the issue is what doctors would actually recommend, then perhaps we should take them into account.

Anyway, the court found that the clinical significance issue was really better addressed in evaluating implicit falsity. I’m not sure I buy that, because I find it hard to imagine why advertisers would ever make claims about statistical significance except to lead consumers to conclusions about clinical significance, which meets the standard for “necessary implication.”

TAP tried to bolster its argument with FDA regulations that state that an ad may be false if it uses the concept of statistical significance to support a claim that hasn’t been demonstrated to have clinical significance. The court rejected this, because the FDCA isn’t privately enforceable. (Again, while plaintiffs do have to show violations of the Lanham Act, not just the FDCA, the FDA’s regulations reflect the idea above – that there’s a necessary implication when an advertiser makes claims of statistical significance, especially to a lay audience – and probably should be usable as evidence in support of this conclusion.)

Because AZ was entitled to summary judgment on the literal falsity claim and the internet implicit falsity claim, all that was left was the TV implicit falsity claim, and of that, only the theory tested in the TV survey, that the ads falsely conveyed the idea that Nexium was better at treating all the symptoms of acid reflux, not just better at healing damage. The court stated in a footnote that TAP couldn’t rely on results from one medium to show falsity in another medium. Without further explanation, this is unconvincing – if the claim is the same, why should it make a difference whether the delivery mechanism was TV, radio, internet, or Pony Express? Perhaps, however, the TV ad copy was sufficiently different to justify this conclusion; otherwise it’s a needless restriction on the ability of plaintiffs to challenge a modern multichannel advertising campaign.

Given that the court already excluded TAP’s damages expert, the only issue remaining is the falsity of a discontinued ad campaign, so the court ordered the parties to conference, presumably with an eye towards dismissal.

No comments: