The New Map of Pharmaceutical Innovation: A Q&A with Michael Scott Kinch

When a drug in development is the subject of litigation or arbitration, determining whether a company’s decisions were reasonable at the time of development often requires evidence about comparable therapies and their geographic origins.

Alongside researchers from Stony Brook University and Purdue University, Analysis Group affiliate Professor Michael Scott Kinch assessed more than 17,000 investigational and approved medicines that entered clinical development between 2000 and 2024. That assessment shows that the number of new drugs moving through the development process each year has grown sharply and that the geographic sources of innovation have broadened well beyond the traditional centers of prior decades. Looking at where those drugs originated, the US led overall. However, by 2023, China had reached a statistical tie with the US in the number of innovative therapies it had developed, a position it held through 2024.

In a Q&A with Managing Principal Jee-Yeon Lehmann, Professor Kinch addressed geopolitical contributions to pharmaceutical innovation, how those contributions may affect litigation and arbitration involving drug development, and whether and how AI could factor into innovation going forward.

How is the increasingly global nature of pharmaceutical innovation changing the way that companies source or develop their next asset?

Michael Scott Kinch: Chief Innovation Officer, Stony Brook University

Certain fast-emerging markets offer a clear example of how quickly a country can build a credible innovation ecosystem when investment, infrastructure, and policy reform align. China, where drug development has grown robustly in the last decade, is one prominent example. The broader lesson from that example – extending past any single country – is that innovation ecosystems can develop quickly. That ability for ecosystems to evolve rapidly is one of many factors companies may weigh as they refine their valuation and licensing approaches over time.

This could have implications for how companies identify their next asset. Understanding the fundamentals of evaluating that asset – including the underlying science, any unmet needs, the competitive landscape, and how a product is different from others on the market – remains central. Many companies may also consider products from a wider set of geographies, earlier on in their life cycles.

Why does a drug’s geographic origin matter in litigation or arbitration involving pharmaceuticals?

As deals increasingly involve pharmaceutical assets that originated outside the traditional US and EU hubs, geographic origin can become a central issue in litigation and arbitration. Assessing a disputed product may depend upon assumptions regarding related development risk, competition, and future market dynamics, all of which could be argued in a specific geographic context. Breach-of-contract claims, licensing and royalty arbitrations, and post-closing M&A litigation, such as earnout and representations-and-warranties claims, can revolve around a central question of whether a company’s decisions about a drug in development were reasonable given what was knowable at the time.

Answering that question typically calls for a careful look at the contemporaneous record – including the scientific data, competitive intelligence, and comparable programs available when decisions were made – rather than an evaluation shaped by hindsight. Factfinders may find it helpful to consider the broader innovation ecosystem in which a therapy was developed to obtain a fuller picture of the information available to decision makers at the time. Development practices, regulatory pathways, and industry norms can differ across regions, and understanding those differences can help explain why a particular R&D decision was made. Identifying where those practices, pathways, and norms align and where they diverge can help developers and investors present a complete and accurate account of their decisions.

How might this affect the way companies assess value and development risk?

Development risk, competitive dynamics, and future market conditions remain key considerations when valuing pharmaceutical assets. Historical benchmarks provide a valuable foundation for such valuations, and they are most informative when applied with attention to the specific characteristics of the asset, since the likelihood of success can vary considerably across therapeutic areas, modalities, and development stages. 

As I explained in a related Analysis Group Q&A on drug development trends, historically, the average probability that a drug entering clinical testing eventually wins FDA [US Food and Drug Administration] approval across all categories is well under one in eight. That figure masks substantial variation. In categories like monoclonal antibodies, for example, the chance of approval has a wide range: something between a one-in-three and a one-in-six chance. And the monoclonal antibodies that have succeeded have spent close to eight years in clinical testing, on average.

Numbers like that are most informative when they are read at the right level of specificity – by modality (a drug’s chemical or biological form), by target (the specific protein or pathway it acts on), or by mechanism of action (how it produces its effect) – rather than as an industry-wide average. Program-specific factors, such as the strength of early clinical data, the validation of the target, and the experience of the development team, can also inform where a particular asset falls within those ranges. The context in which an asset was developed can also be one of several relevant considerations in understanding its value.

Jee-Yeon Lehmann: Managing Principal, Analysis Group

Looking ahead, how do you expect AI to affect this landscape?

If globalization has already reshaped where pharmaceutical innovation happens, AI and computational drug discovery may further influence that trend. For instance, AI may help speed up certain aspects of drug discovery and target identification, although its ultimate impact remains to be seen. Clinical development, regulatory review, manufacturing, and commercialization will continue to require substantial expertise and investment, so the effects of AI are likely to be felt unevenly across the development life cycle. 

Because AI-enabled programs have a limited track record, there is not yet enough history to establish separate benchmarks for their likelihood of success. When such assets are valued or disputed, it will be important to evaluate them on the evidence specific to each program rather than on assumptions about the technology as a whole.

Going forward, investors and industry stakeholders may consider how existing frameworks for evaluating innovation can be adapted to account for AI-enabled approaches. Organizations of all sizes that can adapt to where and how innovation occurs will be well positioned over the next decade. ■