Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2025 Oct;18(10):e70365.
doi: 10.1111/cts.70365.

Predictive Analysis for First Submission of Generic Drug Application for Orphan Drug Products Using Random Survival Forest

Robert Hopefl  1 , Jing Wang  1 , Abhinav Ram Mohan  1 , Wei-Jhe Sun  2 , Myong-Jin Kim  2 , Meng Hu  1 , Lanyan Fang  1
Affiliations

Predictive Analysis for First Submission of Generic Drug Application for Orphan Drug Products Using Random Survival Forest

Robert Hopefl et al. Clin Transl Sci. 2025 Oct.

Abstract

Rare diseases affect a small population of patients, resulting in low incentives for developing orphan drug products (ODPs). The United States Congress passed the Orphan Drug Act of 1983 to incentivize pharmaceutical manufacturers to develop drugs to treat rare diseases. However, ODPs generally have higher treatment costs than non-ODP treatments. Developing generic ODPs can benefit patients by increasing market competition and providing alternate treatment options. This research aims to identify factors influencing the first submission of abbreviated new drug applications (ANDAs) for generic orphan drugs. Data were collected from the U.S. Food and Drug Administration (FDA) databases (including but not limited to the FDA Orphan Drug Designations and Approvals database, Orange Book, and the internal drug product complexity designation) and the IQVIA sales database to inform the drug product information, regulatory factors, and pharmacoeconomic factors. The Random Survival Forest (RSF) machine learning method was employed to conduct the analysis for New Chemical Entities (NCEs) and non-NCEs. The modeling analysis was adequately assessed through both internal and external validations. For NCEs and non-NCEs, the RSF was able to predict ANDA submission with a repeated cross-validation C-index of 0.675 ± 0.0261 (C-index of full training dataset: 0.915) and 0.754 ± 0.0441 (C-index of full training dataset: 0.838), respectively. The variables with the highest importance in the RSF model to predict ANDA submission of NCE ODPs were sales data, while for non-NCEs, regulatory data was the most important (i.e., availability of product-specific guidances (PSGs)). As more data becomes available in the future, the RSF methodology will likely be able to predict ANDA submissions of ODPs more effectively. The model-informed results may be utilized in future intervention strategies to promote ANDA submissions for orphan drugs and to increase the availability of generic ODPs.

Keywords: generic drugs; machine learning; orphan drugs; time‐to‐event analysis.

PubMed Disclaimer

References

    1. United States Code, “Title 21—Food and Drugs,” 2018.
    1. US Food and Drug Administration, “Rare Diseases at FDA,” 2024.
    1. H. Althobaiti, E. Seoane‐Vazquez, L. M. Brown, M. L. Fleming, and R. Rodriguez‐Monguio, “Disentangling the Cost of Orphan Drugs Marketed in the United States,” Healthcare (Basel) 11 (2023): 558.
    1. J. D. Chambers, M. C. Silver, F. C. Berklein, J. T. Cohen, and P. J. Neumann, “Orphan Drugs Offer Larger Health Gains but Less Favorable Cost‐Effectiveness Than Non‐Orphan Drugs,” Journal of General Internal Medicine 35 (2020): 2629–2636.
    1. US Food and Drug Administration, “Generic Drugs: Questions & Answers,” 2021.

MeSH terms