The Vertex / Gilead litigation shows how important it is for pharmaceutical and other life science companies to protect their trade secrets and other highly valuable intellectual property. As I have recently written, companies that are not careful about their trade secrets and other proprietary information can find themselves on the wrong side of a lawsuit or even a criminal referral to the Federal Government. Recent case law confirms not only that pharmaceutical companies are prime targets for trade secret theft (both foreign and domestic), but underscores the need to regularly reassess the adequacy of a company’s firewalls for safeguarding intellectual property, such as trade secrets, and other proprietary information. The risks of not doing so can be severe and can include millions of lost dollars in research and development and an even greater loss in competitive advantage.
http://pharmarisc.com/2013/04/the-importance-of-keeping-trade-secrets/
Confidentiality Laws And Secrecy In Medical Research: Improving Public
Access To Data On Drug Safety
Aaron S. Kesselheim and
Michelle M. Mello
+Author
Affiliations
Division of Pharmacoepidemiology and Pharmacoeconomics, Brigham and Women’s
Hospital, in Boston, Massachusetts.
Aaron Kesselheim (akesselheim@partners.org)
Pharmaceutical manufacturers have long considered results collected from drugs’ clinical trials to be confidential information or trade secrets, even after submission to the Food and Drug Administration (FDA). We describe FDA policies regarding disclosure of clinical trial data and evaluate how courts have interpreted the Freedom of Information Act in cases seeking access to unreleased information. Recent examples of approved drugs later found to have dangerous side effects show the importance of complete dissemination of safety information. We suggest regulatory and legislative policy changes regarding how the FDA handles confidential information that can improve understanding of the risks of prescription drugs....
Protection for “trade secrets” originally arose in common law, or judge-made law, because copyright and patent statutes protecting IP did not adequately cover confidential business information. Copyrights prevent unauthorized reproduction of original works of authorship, but not the ideas behind them.6 Patents cover specific products or processes, not information, and acquiring a patent involves disclosing the formulas underlying discoveries, which in certain cases can weaken a company’s market position.7 To encourage companies to invest in collecting information that could help secure competitive advantages in the marketplace, judges decreed that the law should protect secret commercial information against appropriation by other companies.
In 1939, legal scholars proposed a consensus understanding of what information could be considered a trade secret in an influential treatise called the Restatement of Torts, defining it as “any formula, pattern, device, or compilation of information which is used in one’s business and which gives him an opportunity to obtain an advantage over competitors who do not know or use it.”8 In subsequent years, many state legislatures adopted variations of this broad definition as statutory law. Although the U.S. government never formally adopted a definition of trade secrets, many federal statutes recognized that companies possessed proprietary business information. For example, the Food, Drug, and Cosmetic Act (FDCA) barred FDA employees from disclosing “any method or process which as a trade secret is entitled to protection” acquired in the FDA review process.9 Similar statutes applied to other agencies, and each agency promulgated its own rules about how to treat such information.10
In 1966, however, the Freedom of Information Act (FOIA) altered federal agencies’ confidentiality obligations.11 FOIA created a general rule that the public may request access to information revealing the workings of federal agencies, unless the information falls within one of nine statutory exemptions. Under exemption four, agencies need not release information constituting “trade secrets and commercial or financial information [that is] privileged or confidential.”12 As a result, although nearly all states have broad trade-secret protections, the key question among federal agencies, such as the FDA, is how to balance granting confidentiality to commercial information with the public-disclosure responsibility of FOIA.
Clinical trial data collected by pharmaceutical manufacturers establish the safety and efficacy of their products. This information is commercially valuable because it is costly to acquire and may provide the basis for FDA approval of a product or a change in a company’s strategic direction. As a result, clinical trial data fall within most states’ broad definitions of trade secrets. However, the FDA’s treatment of these data is more complex. Pharmaceutical manufacturers make two formal data submissions to the FDA in connection with new drugs: (1) an Investigational New Drug (IND) application before initiation of human trials, and (2) a New Drug Application (NDA) reporting clinical trial results before marketing. In deciding whether to make these data available to the public, the FDA must balance its obligations to safeguard proprietary information against its FOIA obligations.
Exhibit 1⇓ describes the FDA’s current rules governing public disclosure of data submitted in NDAs. During the premarketing period, the FDA discloses no information from safety-and-efficacy studies, except that patients receiving investigational drugs can obtain safety reports relating to the drug.13 After approval of an NDA, the FDA releases a Summary Basis of Approval, a synopsis of reviews made by FDA staff, including the medical officer and pharmacologist, during the approval process.14 The Summary Basis must contain a “summary of the clinical data section of the application, including the results of statistical analyses of the clinical trials,” but it is not a full report of a drug’s safety and efficacy.15 There is no standard format, and information considered sensitive is redacted. FDA regulations allow the drug manufacturer to draft the Summary Basis, with the FDA’s role limited to reviewing and revising the submission.16Some information can also be released as part of a report from the FDA advisory committee process, as the committee responds to the drug sponsor’s statements.17
EXHIBIT 1
The FDA can make specific data available upon public request in narrow circumstances: If the application is not approved, all other avenues of appeal are exhausted, and no further work is being done on the application.18 Information gathered in support of applications that the FDA does not approve may still be of particular interest in the case of “supplemental NDAs,” where a manufacturer seeks approval for a new use of a drug that is already on the market. Safety data in this case may be of keen interest to the medical community because physicians often prescribe drugs “off-label”—that is, for indications not specifically FDA-approved—before a supplemental NDA is submitted and after it is rejected. Clinical trials in the supplemental NDA can shed more light on an already-approved medication; however, notably, the FDA does not question or evaluate a sponsor’s claim that further work will be done on an unapproved drug.
These rules reflect the FDA’s understanding that research data are entitled to protection as proprietary information. Part of the rationale is to prevent competitors from using data on safety and effectiveness gathered by another manufacturer to obtain approval of generic alternatives or similar drugs within the same class. The FDA originally adopted the broad Restatement approach, considering as a trade secret all nonpublic data submitted in the drug approval process.19 Even after Congress enacted FOIA, the FDA maintained its policy of strict confidentiality until legal challenges from consumer groups, most notably Public Citizen, forced change.20
A series of cases has demarcated the path the FDA treads between its disclosure obligations under FOIA and its confidentiality obligations. In the early 1980s Public Citizen sued under FOIA to obtain postmarket safety and efficacy data on an intraocular lens after the FDA withheld the data as trade secrets. The appeals court disagreed with the FDA’s broad understanding of trade secrets and held that a narrower definition requiring that the agency show “a direct relationship between the information at issue and the productive process” was appropriate in light of the public interests involved in FOIA requests.21 The court concluded that the relationship between clinical trial data and the productive process was “tangential at best.” However, the information might still fall under FOIA exemption four’s provision for confidential commercial information, if disclosure would likely either impair the government’s ability to obtain necessary information in the future (not generally a concern in FDA cases, where certain information is required for marketing approval) or cause substantial competitive harm to the company.22
In 1994 the FDA modified its definition of trade secrets to incorporate the “direct relationship” requirement, and courts have held that clinical trial data do not constitute trade secrets under this narrower standard.23 However, the FDA continued to limit release of submitted data, invoking the commercial-and-confidential-information provision in FOIA to oppose requests for unreleased information based on the prospect of substantial competitive harm.
At the level of both the advisory committee and the full FDA, some data requests have been found not to threaten competitive harm. The Federal Advisory Committee Act requires that the FDA and other government bodies make information presented to their advisory committees open to public review, although it is subject to the same restrictions as FOIA on confidential commercial information.24 After it considered the celecoxib (Celebrex) NDA, the FDA did not disclose certain details of public reports of clinical trials that were presented to the advisory committee supporting Pfizer’s application to market the drug.25 Public Citizen sued to have access to the “full package” of research data, but Pfizer argued that the data would aid its competitors’ drug development and make patients less likely to volunteer for drug trials. However, the court found these claims too vague and ordered the FDA to release the data.26 It reasoned that celecoxib-specific data could not establish the safety or efficacy of a different drug. The case also resulted in a separate settlement in which the FDA agreed to release to the public materials presented at an advisory committee meeting at least twenty-four hours before the meeting. Some details may still be redacted; the FDA does not routinely share all drug safety or efficacy data—including “full reports of raw clinical or preclinical data.”27
Two other cases involved specific data requests to the FDA. In Teich v. FDA,Public Citizen sought results from animal studies and consumer complaints relating to the safety of Dow Corning’s silicone breast implants.28 Although the request specifically excluded product specifications and other manufacturing information, the FDA deemed the information confidential. Dow Corning argued that disclosure could facilitate competitors’ safety testing. The court disagreed, characterizing the competitive harm as “negligible” and disclosure as “unquestionably in the public interest.” In another case, Public Citizen requested a postmarketing study protocol relating to Bristol-Myers Squibb’s glucophage (Metformin). The court ordered the FDA to release the information, finding that competitors were unlikely to gain an advantage from the protocol.29
On the other hand, in some notable instances, courts have supported the FDA’s refusal to disclose information. Citizens Commission on Human Rights v. FDAupheld the FDA’s power to withhold all records pertaining to a supplemental NDA for a new use of Lilly’s antidepressant fluoxetine (Prozac).30 The court noted that data in supplemental applications are especially sensitive from a competitive standpoint because they relate to new market strategies. It also determined that the potential public impact of such data was negligible, because “it is only if these supplements are approved that the public has an interest in the information.”
Similarly, a 1999 case stemmed from Public Citizen’s request for “all documents concerning pre-clinical and clinical studies” relating to pharmaceutical products that Schering had abandoned because of safety issues after submitting an IND application.31 Schering argued that the toxicity studies might be useful to competitors because the results suggested that the drug worked through a particular mechanism. The court accepted this argument, holding that disclosure of the data at the IND stage could help competitors bring rival drugs to market. The court made its decision exclusively on the likelihood of competitive harm, largely rejecting Teich to the extent that it considered the “public interest side” of FOIA disclosure. Yet the court required little specificity from the manufacturer or the FDA about the nature of the competitive harm.
Collectively, these cases indicate that although courts will not permit clinical and pre-clinical trial data to qualify as a trade secret, some types of data can qualify for exemption from FOIA disclosure as confidential commercial information. Such a finding is more likely where the scope of the data request is broad and where the drug manufacturer can describe with some specificity how a competitor could use the information to expedite the development and approval of a rival drug. Presenting the data before an advisory committee may open the information up to disclosure as well.
These cases also demonstrate why FOIA requests are a suboptimal vehicle for disseminating safety data. FOIA requests are generally the only avenue available to consumer groups, researchers, and physicians seeking to access information not released by the FDA.32 If the FDA and manufacturers resist the FOIA request, litigation—invariably protracted and expensive—becomes necessary. If the issue reaches judicial review, the petitioner may well find the court unwilling to consider the public’s interest in such information in determining what qualifies as confidential commercial information.
The potential public health implications of limiting access to clinical trial data already submitted to the government have become increasingly clear. Undisclosed research data may reveal dangerous drug side effects that call into question manufacturers’ claims or the FDA’s decisions. A number of recent examples of approved drugs later discovered to have major safety risks illustrate the value making such data available for independent analysis.
In 2001 the FDA approved nesiritide (Natrecor) for treatment of congestive heart failure, despite data hinting that the drug was associated with impaired renal function and increased mortality relative to other intravenous therapies.33Although the FDA requested that the manufacturer, Scios, further study the incidence of such side effects, it did not delay nesiritide’s approval. Four years later, researchers not affiliated with the FDA used meta-analyses of the summaries of initial data and data published after its approval to show that the side effects and mortality rates were statistically significant and clinically troubling.34 Scios and the FDA later issued a physician advisory recommending that nesiritide’s use be restricted.35
Another example occurred in September 2005 when an FDA advisory committee recommended approval of muraglitazar (Pargluva) for the treatment of diabetes and dys-lipidemia. Soon after, other researchers reanalyzed the summary data released by the FDA and found statistically significant evidence of increased risk of cardiovascular complications and mortality.36 Bristol-Myers Squibb removed the drug from FDA consideration and soon ended research on it entirely, citing the fact that the supplemental studies needed to better define the cardiovascular side effects could take up to five years.37
For both nesiritide and muraglitazar, a more comprehensive analysis could have been conducted had the analysts had access to the full set of data submitted to the FDA. This point is illustrated by controversies over reporting of studies of the anti-inflammatory medications valdecoxib (Bextra) and rofecoxib (Vioxx). In 2001 Pfizer submitted valdecoxib to the FDA as a treatment for symptoms associated with osteoarthritis, rheumatoid arthritis, dysmenorrhea, and acute pain. The FDA approved the drug for the first three indications. Pfizer successfully petitioned the FDA to restrict clinical trial data related to studies for the acute pain indication, which included the results from a clinical trial described as the “only substantial multidose safety database.”38 It was not until Public Citizen initiated a lawsuit in 2004 that the pertinent data were released, revealing the FDA’s review that these safety data demonstrated “an excess of serious adverse events including death…warrant[ing] further investigation before valdecoxib can be considered safe and effective for the treatment of pain, particularly multidose therapy in the perioperative setting.”39 In April 2005 valdecoxib was withdrawn from the market when additional studies revealed an increased incidence of postoperative acute myocardial infarction.40
A 2005 editorial in the New England Journal of Medicine suggested that the authors from a pivotal study of rofecoxib also might have omitted relevant data.41 The manufacturer, Merck, responded that it submitted complete data to the FDA.42 However, neither journal reviewers nor independent researchers were able to fully evaluate the completeness of the article or the safety of the drug.
The FDA’s disclosure policies are particularly troublesome in the context of supplemental NDAs. Clinical trials undertaken to support supplemental NDAs may reveal safety risks that affect patients who are prescribed the drug off-label, yet the data are inaccessible unless the application is denied and abandoned. The court in the Citizens Commission case saw no public interest in information about off-label uses, but the prevalence of off-label prescribing belies this view.
For example, Lilly submitted a supplemental NDA in 2002 proposing that duloxetine (Cymbalta), an antidepressant, be approved to treat stress urinary incontinence. It later withdrew the application because clinical trials detected an elevated risk of suicidality.43 In 2005 the FDA posted an advisory warning that anti-depressants, including duloxetine, posed this risk.44 However, the study data reportedly have remained unavailable (perhaps because the company is continuing to work on this use of duloxetine). When her FOIA request seeking this information was turned down, a journalist was told by an FDA official, “My hands are tied.”45 Indeed, Lilly objected that even posting the FDA warning violated confidentiality rules.46
When GlaxoSmithKline studied the use of its antidepressant paroxetine (Paxil) in treating major mood disorders in pediatric patients, it found indications that the drug might have increased the risk of suicidal thoughts in this population. In 2004 the New York attorney general charged that GlaxoSmithKline had fraudulently concealed adverse findings from two studies of the proposed use.47The company pointed out that complete data had been submitted to the FDA, in the context of the manufacturer seeking six extra months of patent extension granted by federal legislation for performing such clinical trials in pediatric populations. However, the FDA’s policy prohibited public disclosure of the information, so there was no mechanism for informing off-label prescribers of the risk....
Aaron Kesselheim (akesselheim@partners.org) is a clinical fellow in medicine, Division of Pharmacoepidemiology and Pharmacoeconomics, Brigham and Women’s Hospital, in Boston, Massachusetts. Michelle Mello is the C. Boyden Gray Associate Professor of Health Policy and Law at the Harvard School of Public Health, also in Boston.