If a drug manufacturer pays the publication fee, is an open access article an ad?

This case doesn’t resolve the issue, but it’s an interesting one to contemplate.

Ony, Inc. v. Cornerstone Therapeutics, Inc., 2012 WL 1835671 (W.D.N.Y.)

Ony sued over defendants’ publication of an article containing allegedly false and misleading statements about its product, an animal derived surfactant known as Infasurf, which is used to treat Respiratory Distress Syndrome (“RDS”) in premature infants.  Such infants often suffer from inadequate lung development, including naturally produced surfactants.  The parties compete with animal-derived surfactants.

The complaint alleged that Chiesi commissioned a study to support the claim that Curosurf was superior.  Chiesi hired defendant Premier Research Services to “provide a database to support the desired conclusion,” and hired three defendant-authors (Ramananthan, Bhatia, and Sekar) to submit the findings to pediatric medical societies.  The authors and Ernst, a Premier employee, submitted a study for publication in the Journal of Perinatology that included allegedly false claims that Curosurf-treated infants had a lower mortality rate than Infasurf-treated infants, e.g., “Result: Calfactant [Infasurf] was associated with a 49.6% greater likelihood of death than poractant alfa [Curosurf].”  The article claimed that this relationship held up even adjusting for patient and hospital characteristics.  The complaint further alleged that the Journal (whose parent was also a defendant) was perceived as a reliable source by neonatologists.

Ony alleged that the data used for the article weren’t based on an actual clinical study, but rather from various reporting hospitals and doctors, making the study retrospective and “subject to selective distortion.”  Specifically, the mortality data were unreliable because of deliberate omission of length of stay data, since length of stay is inversely proportionate to mortality.  Allegedly, if the length of stay data had been included, it would have been obvious that the differences in results came from differences in patients treated.  Ony further alleged that the authors didn’t cite at least one article contradicting their findings, despite knowing about it, and tha the defendants submitted the article for publication knowing that it was deficient.  In fact, Ony alleged, it didn’t think the doctor-authors actually wrote the article; some other agents of the manufacturer/seller did.

Premier allegedly benefited economically from submitting and publishing the article, which “served as advertising for Premier's database and related services, and announced to pharmaceutical companies and other providers of medical products and services ... that the engagement of Premier in connection with similar self-serving ‘studies' would be to their economic benefit as well.” 

One of the two reviewers found the article’s conclusions unreliable, but the journal published it anyway.  Ony alleged that this was based on part on the fact that Bhatia was an associate editor at the journal and Sekar was on the editorial board.  The manufacturer/distributor defendants then allegedly paid the publication fee for the article, which enabled it to come out in open access format available to nonsubscribers.  They then issued a press release touting the article and began distributing it to current and potential customers.  Because the journal refused to retract the article at Ony’s request, Only alleged that the journal and its publisher the American Academy of Pediatrics “are sanctioning and condoning scientifically unreliable information which influences purchasing decisions by hospitals and prescribing decisions by neonatologists.”

Ony therefore alleged that publication and distribution of the article violated the Lanham Act, NY GBL § 349, and constituted tortious interference as to the manufacturer/distributor, and also constituted injurious falsehood as to every defendant.

The authors, including Ernst, contested personal jurisdiction.  None of them were domiciled in NY, and the article was written and researched outside of NY.  Publication was not negotiated within NY; publication was not negotiated at all but rather conditioned on peer review and left to the editorial discretion of the defendant editor.  The authors weren’t paid.  Ony argued that defendant Nature Publishing Group requires all authors to sign a license prior to submitting any papers for publication, which contains a choice of law provision choosing NY law and submitting to the non-exclusive jurisdiction of the courts of NY.  But choice of law provisions are insufficient on their own to confer jurisdiction, and a forum selection clause is meaningless without express consent to personal jurisdiction.  The agreement language referred only to “non-exclusive jurisdiction” and had no express consent to personal jurisdiction.

The claim against the authors sounded in defamation, and courts considering NY’s long-arm statute have concluded that something more than the distribution of a defamatory statement within the state is required to establish long-arm jurisdiction.  For Ernst and Ramanthan, the only New York contact alleged was the submission of the article itself, and that wasn’t enough. What about Bhatia and Sekar as participants in the journal’s editorial activities?  An employee’s contacts with a forum state are assessed individually rather than according to his/her employer’s activities.  Their duties were limited to reviewing articles emailed to them and returning comments.  This was also insufficient to justify personal jurisdiction in NY.

Under the facts, defendants’ submission of the article to a NY publisher was akin to a mere solicitation or placement of an ad, which is insufficient to constitute doing business in NY.  There was no personal jurisdiction over the four author-defendants.

The court then turned to the motion to dismiss for failure to state a claim.  Statements of opinion can’t support claims under the Lanham Act, GBL § 349, or injurious falsehood/defamation.  This includes a hypothesis or opinion offered after a full recitation of the facts on which it is based.  The claims here were based on an academic article in a scientific journal, and academic freedom is a special concern of the First Amendment.  Still, there’s no license to make false statements.  The fact/opinion divide can be assessed by considering (1) whether the specific language in issue has a precise meaning that is readily understood; (2) whether the statements are capable of being proven true or false; and (3) whether either the full context of the communication in which the statement appears or the broader social context and surrounding circumstances are such as to signal that the statement is opinion or fact.  Could a reasonable reader have concluded that the article was conveying facts about the efficacy of Ony’s product?

Ony pointed to the article’s conclusions that calfactant (Infasurf) was associated “with a 49.6% greater likelihood of death” or a “significantly greater likelihood of death” than poractant alfa (Curosurf).  But Ony didn’t allege that calfactant was in fact more effective.  (Not clear why that’s the standard, as opposed to “not significantly less effective.)  Instead, it alleged that the article’s conclusions were unreliable and therefore misleading.  Thus, the court framed the issue as whether the article sufficiently stated the facts on which its conclusions were based or impermissibly implied that the conclusions were supported by additional undisclosed information. 

Commentary: this is a defamation analysis, but applying it in this way is completely inconsistent with the Lanham Act caselaw about establishment/“tests prove” claims, where showing unreliability of the tests used to make the assertion is a well-recognized path to liability.  Had Ony picked its defendants better, and brought only Lanham Act claims against a competitor, the court might not have been led to apply this standard.  While that would still leave the article out there even if Ony fully succeeded, it was never very likely that Ony would completely suppress the article, and more reasonable to imagine that Ony might be able to stop its use in advertising and promotion.

Once the court applied this standard, Ony’s claim was doomed.  The peer-reviewed journal at issue was directed to a highly specialized group familiar with the issues.  The article contained an initial section detailing the patient data and research methods used.  Though it didn’t mention length of stay data, it did disclose the patient criteria considered.  Thus, there was no implication of undisclosed facts.  The context also supported the conclusion that an average reader would perceive the statements as “debatable hypotheses” rather than “assertions of unassailable fact.”  The authors acknowledged that the retrospective nature of the study posed limitations and that there could be additional factors affecting the stated conclusions.  They specifically stated that the findings prompted an inquiry into the reasons for the outcomes, and that “[t]he most likely explanation may be due to different surfactant doses administered to the infants included in the database,” as doses of poractant alfa tended to be twice as large as doses of the other two surfactants. Thus, the court couldn’t conclude that a reasonable reader would have received a message that the article was conveying proven facts about the efficacy of Ony’s product.  “[A]ny perceived fault in the method by which the authors reached their conclusions should be subjected to peer review rather than judicial review.”

Ony alleged that the manufacturer/distributor defendants paid Premier and the authors to compile data to support a favorable conclusion, but that was clearly indicated in the article under the heading “Conflict of Interest.”  There, the authors acknolwedged that the “study was sponsored by Chiesi Farmaceutici SpA, the manufacturer of poractant alfa. Frank R. Ernst is an employee of Premier, which contracted with Chiesi Farmaceutici SpA to conduct the study. Rangasamy Ramanathan, Kris Sekar and Jatinder Bhatia have served as consultants to Ciesi Farmaceutici SpA.”  This clearly signalled that the authors weren’t disinterested, leaving a credibility assessment to the reader and reinforcing that the conclusions were opinion.

The tortious interference claims also necessarily failed, even after Ony identified specific hospitals whose relations were allegedly interfered with; a defendant can advance its own economic interests unless it employs wrongful means.  Since promoting and distributing the article wasn’t misleading, there was no wrongful conduct alleged.