Modeling Drug Development: Improvements Still Needed

The Congressional Budget Office (CBO) informs Congress about the costs or revenue changes associated with proposed legislation, a challenging enough task on its own. Sometimes, the CBO is asked to estimate effects beyond budget impact such as the implications of centralized administrative pricing on drug development. Thus, the CBO has worked to establish and improve its modeling capabilities for assessment.  A 2021 CBO working paper described a flexible simulation model, subsequently in 2023 CBO solicited input, received feedback from Congress, and in 2024 released an updated model.

In this model, a representative firm decides whether to invest at each phase of clinical development. The simulation takes place across a range of expected revenues, costs, and drug characteristics such as competition, mode of administration, and whether the drug is a large or small molecule. Modeled decisions are based on historical revenues, incorporating estimated discounts and historical rates of new candidate entry and success at different phases of development. If projected profitability exceeds zero, the representative firm is expected to advance the drug to the next phase of development. A prior phase informs the outcome of the next, reflecting market realities.

While it represents an advancement in methods, we consider attributes in drug development that are not well described in the CBO model, including post-market drug development, capital mobility, and health. It should be understood that this model is currently limited in its ability to predict policy impacts on changes in product development and American life sciences leadership. 

Post Approval Research

The CBO model only considers impacts on new drug development, not the essential role of continued development of drugs post initial approval.

Post approval research is a key part of the drug development lifecycle, with many products continuing clinical development long after initial FDA approval to establish new indications, efficacy in different disease states,  in a distinct population (e.g., children), or reformulations.  Drugs for cardiovascular and antithrombotic indications have half of new indications from post-approval clinical studies, while oncology products are typically tested first on small populations with expansion to more indications and earlier disease states.

Consider lisinopril, approved for the treatment of hypertension in 1988, with subsequent development through the GISSI-3 trial (1994) resulting in an indication for acute myocardial infarction, the 1999 ATLAS trial supporting an indication for heart failure and finally the development of an oral solution (2016) expanding access to millions with swallowing difficulties (e.g.  prior stroke), a neurodegenerative disorder, or who are dependent upon alternative routes for feeding and medication administration  (e.g. with a gastrostomy tube).

Capital Mobility and Global Development

The CBO model also assumes that if a drug candidate is expected to be profitable, even if it is less profitable after a policy change, it will still receive investment for clinical trials to receive approval. Market realities suggest otherwise.

Much early-stage drug development occurs in smaller firms that rely on private or venture capital for initial clinical development. Venture capital is mobile and flows to areas with the greatest expected profitability, taking risk into account. When policy changes affect expected returns on capital, investors will reallocate funds to therapeutic areas with higher expected returns (e.g. metabolic disease) or other potential investments yielding significant returns (e.g. Artificial Intelligence). 

Other countries, including China, represent investment opportunities in a global arms race for pharmaceutical product development, threatening American innovation primacy. China introduced accelerated regulatory approval in 2016, receiving 774 applications in the subsequent two years. The market capitalization of Chinese biopharmaceutical companies subsequently grew from $3 billion in 2016 to $380 billion in 2021 with 23 IPOs in 2020, including 7 of the 10 largest worldwide. The Chinese government is targeting a 2.5% GDP spend on R&D, and seeks to link its academic researchers and industry as part of its Fourteenth Five-Year Plan, raising global competition and national security concerns with China on track to soon conduct more clinical trials.

Health Impact 

Finally, the CBO model does not measure any impact on health. The model calculates the number of new drugs that are not developed. However, if a policy deters investment in drugs with limited health benefits, it may effectively steward scarce resources. Conversely, if a policy deters investment in the clinical study of highly effective treatments for diseases without effective therapies, the negative impact is much greater.

Conclusion

Biopharmaceutical innovation has contributed meaningfully to longer and better lifespans. It is important to recognize the limitations in measures of impact assessments on pharmaceutical markets.  Current models fall short, lacking assessment of post approval research, capital mobility in a global world, and on health.  With pharmaceutical innovation representing one of the bright spots for positive change in health care, appreciating the impact of policy is key to a healthier and more competitive America.