“From Target Discovery to Clinical Drug Development With Human Genetics”, 2023-08-23 ():
The substantial investments in human genetics and genomics made over the past 3 decades were anticipated to result in many innovative therapies.
Here we investigate the extent to which these expectations have been met, excluding cancer treatments.
In our search, we identified 40 germline genetic observations that led directly to new targets and subsequently to novel approved therapies for 36 rare and 4 common conditions. The median time between genetic target discovery and drug approval was 25 years. Most of the genetically driven therapies for rare diseases compensate for disease-causing loss-of-function mutations.
The therapies approved for common conditions are all inhibitors designed to pharmacologically mimic the natural, disease-protective effects of rare loss-of-function variants. Large biobank-based genetic studies have the power to identify and validate a large number of new drug targets.
Genetics can also assist in the clinical development phase of drugs—for example, by selecting individuals who are most likely to respond to investigational therapies. This approach to drug development requires investments into large, diverse cohorts of deeply phenotyped individuals with appropriate consent for genetically assisted trials.
A robust framework that facilitates responsible, sustainable benefit sharing will be required to capture the full potential of human genetics and genomics and bring effective and safe innovative therapies to patients quickly.
…We applied these criteria to data from OpenTargets24, DrugBank25 and the FDA (Supplementary Figure 1). In total, we identified 2,832 FDA/EMA-approved therapies. After exclusion of antineoplastic drugs (n = 277), anti-infectives and antiparasitics (n = 392), hormonal preparations (n = 154), vitamins and analogues (n = 90) and drugs whose target is unknown (n = 902), our search resulted in 1,031 drugs with matching protein-coding target genes. Most drugs (766 out of 1,031 (74%)) acted through multiple targets; in addition, many drugs had more than one indication (794 out of 1,031 (77%)), resulting in 6,690 drug-target-gene indication triplets (Supplementary Table 1). Some direct genetic evidence derived from 8 data sources was reported for 619 triplets (corresponding to 346 different drugs) (Supplementary Note 1). For 98 triplets (corresponding to 80 drugs), such evidence had been reported more than 5 years before drug approval. Manual, literature-based curation documented that genetic information had been essential for 60 drugs—that is, these drugs would probably not have been developed had the genetic association not been discovered. After grouping the drugs by class (small molecule, biological or gene therapy), our analyses identified 47 first-in-class (and 13 follower) therapies for 40 targets that met our definition of being genetically driven (Supplementary Tables 1–3).
These particular criteria captured only a fraction of the impact of genetics on drug discovery and development. We acknowledge, for instance, that we missed drugs such as mavacamten (approved by the FDA in 2022), which was absent from the databases that we used despite the fact that its development depended in part on the discovery of mutations in several genes responsible for hypertrophic cardiomyopathy26,27,28. Nonetheless, we make several observations based on our analysis. The 40 targets correspond to around 6% of the ~500 drug target genes of FDA-approved non-cancer drugs29; the remaining 94% were probably identified using conventional pharmacology, biochemistry or molecular biology approaches. The therapies targeting these 40 gene products were approved for chronic, as opposed to acute, conditions. Most of them (26⁄40) were targeted for treatments of metabolic disorders, whereas the remaining 14 genes were targeted to treat diseases in 7 other therapeutic areas (Figure 2a).