I am a Senior Partner in McKinsey & Company’s Life Sciences, Healthcare and Private Capital practices, based in Boston. You can find my McKinsey website here and my LinkedIn page here. This is my personal website. The views expressed here are my own and do not necessarily represent those of McKinsey.
I created this page to share insights and commentary related to some of my recently published work that analyzes data across the life sciences and healthcare ecosystem and identifies insights for companies and investors.
My interests sit at the intersection of science, technology, business, and investing, and for the past 20 years at McKinsey, I’ve helped clients navigate exactly that intersection, guiding them through the capital-intensive, long-duration, and risk-laden process of discovering, developing, and commercializing new drugs. My passion for science took root at the University of Cambridge, where I earned a PhD in biological physics as part of the inaugural class of NIH-Oxford-Cambridge Scholars. My interest in technology, business, and investing grew at the University of Pennsylvania, where I was an undergraduate in the Management and Technology Program, pursued graduate studies in the School of Engineering and Applied Science, and earned an MBA from the Wharton School. Between Cambridge and McKinsey, I served as a postdoctoral research fellow at the National Institutes of Health, researching protein folding, misfolding, and aggregation.
Commentary
Nature Reviews Drug Discovery · Vol. 22 · August 2023
Herding in the Drug Development Pipeline
1 Overview
The idea that pharma pipelines over-index on validated biology is well established, under labels such as “me-too,” “crowding,” and “lemmings.” What was missing was a quantitative picture. This paper coined the term “target herding” and measured it, using a time-series analysis of pipeline herding across all targets and indications that separates oncology from non-oncology trends and compares the whole industry with the pipelines of the top 10 pharma companies. Herding turned out to be broad-based and not confined to headline targets such as PD(L)1, CD19, and BCMA: it also shows up in rare monogenic diseases, where nine diseases have ten or more assets in clinical development, and fully enrolling all active trials for three of them (thalassaemia, Duchenne muscular dystrophy, and Huntington disease) would require more than 10% of the current US patient population.
We also provided a framework for the factors driving herding and for what it means for pharma strategy. On the drivers, some expand the number of assets pursuing a target (e.g., modality advances aimed at best-in-class assets for validated targets, as in the “battle for BCMA” among cell therapies, bispecifics, and antibody–drug conjugates; and new cell and gene therapy platforms that use well-validated targets to de-risk the core technology), while others constrain the number of targets (e.g., the limited pool of compellingly validated biology, and unsolved technical challenges such as tissue delivery for gene therapies and the immunosuppressive tumour microenvironment that has held cell therapies back in solid tumours).
2 Key findings
We showed that target herding is accelerating, with important implications for biopharma pipelines. The average number of assets per actively investigated clinical target has more than doubled since 2000, from 3 to 7 overall, and from 1.8 to 9 in oncology, a greater than five-fold increase. In oncology, the number of assets has grown 7.6% per year while the number of targets has grown just 3.4% per year, so pipelines are expanding much faster than the biology they are built on. The effect is strongest at the top: the five most active targets in any given year saw a 4.7-fold increase in assets per target between 2000 and 2020. For the top 10 pharma companies, 68% of pipeline assets in 2020 were directed at herded targets (more than five assets per target), up from 16% in 2000 and well above the 30% seen across all companies. This raises fundamental questions about the efficient use of R&D resources and about the differentiation bar required for late-entrant assets.
3 Key figures