The rising cost of new drugs has provoked economic and political dilemmas for more than 60 years now, as the promise of innovative biopharmaceutical products has been tethered to price tags that drift iteratively further and further out of the reach of many Americans. But in recent years, the increasing cost of old drugs has become a vexing problem as well, and one that may prove harder to solve. In the past decade, Americans saved nearly $1.7 trillion dollars on equivalent but cheaper generic copies of off-patent drugs. Yet over the same time period it has also become apparent that specialty biological drugs with the highest price tags (many more than $300,000 a year with some drugs anticipated to break the $1 million dollar barrier in just a few years) will see no simple generic price solution once off patent.^{Reference Herber1} As, Diependaele, Cockbain, and Sterckx discuss in this issue,^{Reference Diependaele2} current pathways for biosimilar, biointerchangeable and other forms of follow-on biological drug approvals approved since the passage of the Biologics Price Competition and Innovation Act of 2009 have not yet and are not likely soon to provide the substantial cost-savings that small molecule generics have provided in the past.

The disappointment of the biosimilar as a solution to rising drug prices can be seen in the dashed hopes that follow-on copies of analog insulin products might bring price relief to patients living with diabetes. Insulin prices have increased exponentially over the last decade, and while the exact reasons for this increase are not clear, observers have noted that lack of robust competition must play some role. 99% of the global insulin market, and 100% of the U.S. market, is dominated by brand-name biological products from just three producers: Eli Lilly, Novo Nordisk, and Sanofi-Aventis, which have increased prices steadily apace in this time period even as their patents expired.^{Reference Beran, Hirsch and Yudkin3} When the first follow-on insulin product was approved for U.S. use in 2015, many hoped that a robust biosimilar market might soon bring insulin prices back down again.^{Reference Heinemann4} Yet both of the two off-patent insulin analog products approved for use in the U.S. (Sanofi's Admelog and Lilly's Basaglar) have been priced almost as high as the brand-name versions. Moreover, both are products of the same three firms: to date, follow-on insulin products have not expanded the American insulin marked beyond an oligopoly.

Diependaele, Cockbain, and Sterckx argue that these and other disappointing results of biosimilar markets result from definitions of intellectual property built into the U.S. biosimilar approval pathway, which allows generous definition of trade secrets and data exclusivities that raise substantial barriers to would-be entrants in the off-patent biological market. It is almost as if this system was designed so that successful biosimilar products would only come from the limited set of firms that already produce brand-name biopharmaceutical products in the first place. This was not the case for the early generic drug industry, which was made up of a very different set of would-be pharmaceutical manufacturers, and which employed very different pricing and marketing strategies compared to brand-name firms.^{Reference Greene5} While it is early days still for the biosimilar market, and it is possible that a truly competitive sector might yet emerge, evidence to date suggests that unless the structure of the approval process is substantially changed, cost savings of biosimi-lars will continue to disappoint.

The authors argue that biosimilars will only yield competitive markets for off-patent biologicals if and when the barriers for development do not require costly (and, as they argue, potentially unethical) efforts to reverse-engineer cell lines and other biological production techniques and prove through clinical trials that large-molecule therapeutics are similar enough to be interchangeable with innovator products. Current interpretations of intellectual property law and data exclusivity arrangements in the biotech sector protect knowledge of these cell lines and other techniques as trade secrets. Yet the boundary between knowledge required to be disclosed to receive a patent or other market exclusivity and knowledge protected by trade secret is far from a fixed thing, and can be changed by executive rulemaking, judicial ruling, or amendment to the existing law. Expanding these disclosures to include key knowledge about the cell lines required to produce these drugs could simultaneously reward incentive for innovation in the biopharmaceutical sector while also enabling a pathway for true generic competition to emerge once these market exclusivities end. From legal and economic standpoints, this proposal has merit and should be strongly considered. From a political standpoint, however, it will almost certainly be opposed by the full force of the pharmaceutical and biotech industries.

This is not, however, the first time that the rights of innovator pharmaceuticals to claim forms of data exclusivity as a means of forestalling generic competition has threatened the ability of a nascent generic drug industry to provide high-quality safe and effective products that compete at a lower price. Just as the 1984 Hatch-Waxman Act — which helped establish the Abbreviated New Drug Application (ANDA) pathway that enabled generic competitors to easily prove therapeutic interchangeability in the small-molecule generic sector — was being negotiated, the case of Roche v. Bolar threatened to raise an analogous barrier to data exclusivity to undermine nascent generic competition. In that case, Roche argued that the use of its brand-name product in trials to demonstrate the equivalence of Bolar's would-be generic version constituted patent infringement, and the Court of Appeals for the Second Circuit agreed with this interpretation of the Patent Act. In this case, the legislature had the political will to make sure that a broad scope of intellectual property would not to continue to present a barrier to entry for generic competition, so the so-called Bolar Amendment was included in the Hatch-Waxman Act of 1984 that clarified the statute and overturned the Second Circuit's interpretation.^{Reference Engelberg6} An amendment to the BPCIA, correcting the continued problem that cell line privacy and data exclusivity plays in raising barriers to biosimilar competition, could yield a market of biosimilars in the early twenty-first century as robust as small-molecule generics became in the late twentieth century.

The remedy suggested by Diependaele and colleagues — mandatory disclosure of original biological cell line — raises broad questions about which parts of intellectual property remain private, and which become public, once the federal government grants effective monopoly power to an innovator in exchange for commercializing that innovation. Reflecting on the differences between the successes of generic drugs since 1984 and the disappointments of biosimilar drugs since 2009, one is left with a sense that the difference between small and large molecule are not merely differences of molecular scale. They are also differences in the political will within American legislative bodies and regulatory agencies to redraw the line between private and public intellectual property in ways that reward true innovation and promote competition. In the present context, when much is being said but little being done to address the soaring price of on- and off-patent drugs, perhaps their suggestion might help political will find a way.