Technology profiles in FDA-approved new molecular entities, 2013-2024: a latent class analysis

Pharmaceutical innovation is often described using isolated approval-time attributes such as modality, first-in-class status, or expedited designations, but this approach does not reveal how multiple technological attributes are combined within the approved-drug portfolio.

We analyzed 496 therapeutic new molecular entities approved by the U.S. Food and Drug Administration between 2013 and 2024 using therapeutic modality, mechanism of action, target class, and therapeutic area. Latent class analysis identified recurring technology profiles, and their prevalence was compared across four three-year approval periods using Pearson's chi-square test.

A retained four-class solution identified an enzyme-centered inhibitory small-molecule profile (35.3%), a receptor-centered mixed-mechanism profile (27.8%), an antibody receptor-ligand profile (22.4%), and a heterogeneous alternative-modality profile (14.5%). The inhibitory small-molecule profile was the largest in all four periods, while all four profiles were represented throughout the study period. The antibody receptor-ligand profile accounted for 15.2% of approvals in 2013-2015 and 26.1% in 2022-2024, whereas the receptor-centered mixed-mechanism profile accounted for 34.3% and 24.6%, respectively. The overall profile distribution was not significantly associated with approval period (χ² = 7.94, p = 0.541).

FDA-approved NMEs were organized into four recurring multidimensional technology profiles. The enzyme-centered inhibitory small-molecule profile remained the largest, while antibody, receptor-centered, and alternative-modality profiles also represented substantial portfolio components. These findings demonstrate the coexistence of multiple technological patterns and show how approval-time attributes are combined in approved medicines.