Beyond Animal Testing: How FDA Modernization Act 2.0 Could Transform Retinal Drug Development
How the shift away from animal testing is reshaping the path from lab to clinic
It is no longer a distant future where in vitro human cell-based models dominate the testing market for drug discovery and development. Not forgetting the elephant in the room—artificial intelligence (AI), the use of computer simulations and machine learning (ML) for in silico models to predict drug behaviour, toxicity, and efficacy is also now a competitive necessity for the drug industry rather than a future possibility. The archaic Federal Food, Drug, and Cosmetic (FD&C) Act requiring animal testing for drug approval is finally being challenged with the passing of the U.S. Food and Drug Administration (FDA) Modernisation Act 2.0 in December 2022, presenting a key driving force for this paradigm shift away from animal-based testing. This legislation marks a pivotal moment for drug development, including retinal therapeutics, by potentially enhancing translation success and accelerating innovations.
While this regulatory reform is a huge hoorah for animal rights activists and vegan communities, there are considerable hurdles to be overcome for this to be a reality. Significant improvements must be made to these in-vitro assays to capture disease mechanisms accurately, coupled with rigorous validations on in silico models to allow drug testing in a reliable and scalable manner. Apart from more groundwork in the science, infrastructural changes and expansion to funding and training in non-animal approaches are necessary to support this shift.
Revisiting the Eight Decades of Animal Testing Mandates
The 1938 Federal FD&C Act was enacted to ensure product safety following a fatal pharmaceutical incident. While the Act did not mandate animal testing, it required manufacturers to demonstrate product safety before marketing. Animal testing emerged and was quickly normalized as the primary method to meet this requirement. The evolution of regulatory science and persistent reliance on animal models consequently emanated from institutional frameworks, established routines and cultural norms that have long shaped scientific and industry practices in drug development.
Over the recent decades, there have been growing recognition of limitations in translating animal data to human outcomes that contributed to the notorious valley of death in drug discovery. The pharmaceutical industry spends an estimate of $28 billion annually on preclinical research with a staggering 93% failure rate – only 6-7% of promising laboratory candidates reach market approval1-3. Despite the breakthrough of anti-VEGF therapies in 2004 and the success of several blockbuster drugs that followed, ophthalmic drugs with positive Phase 2 results have only had about a 47% chance of advancing to approval, lagging behind success rates in many other therapeutic areas4. The reasons are consistent across medicine: 60% of clinical trials fail due to lack of efficacy and 30% from unexpected toxicity, revealing a fundamental disconnect between current testing models and human physiology. Beyond financial losses, this results in delayed treatments for patients, diminished investor confidence, and misallocated research resources. The FDA Modernization Act 2.0 offers a solution by authorizing human-relevant alternatives.
Understanding FDA Modernization Act 2.0
The new law5,6 allows drug companies to use modern human-based research technologies instead of animal testing in preclinical drug trials when seeking FDA approval. These alternatives include in vitro human cell-based assays, such as patient-derived cells, stem cell-based models, 3D cell culture, organoids, organs-on-chips and bioprinting of human tissues, and in silico computer models. Further, the Act eliminates redundant animal testing for biosimilar drugs, generic versions of existing biologics.
Check out our previous editorial piece on Patent Cliffs, where we showcased the emergence of biosimilars in retinal therapeutics.
Following FDA’s footsteps, the National Institutes of Health (NIH) also recently announced their plans7 to establish the new Office of Research Innovation, Validation, and Application (ORIVA) to prioritize the development, validation and scaling of non-animal methods for biomedical research.
The Retinal Therapeutic Landscape: Current Bottlenecks in Preclinical Studies
The FDA Modernization Act’s move to phase out mandatory animal testing for regulatory processes, particularly for monoclonal antibodies and other complex therapies8, is a progressive step. This new transition to non-animal testing does not compromise safety standards; it enhances them by using human-relevant methods that more accurately reflect how drugs work in people. This is particularly important for complex conditions like eye diseases, where animal models often fail to predict human responses. Common in vivo models for eye research – including vertebrates such as rodents, rabbits, primates, pigs, felines, canines, and zebrafish, and invertebrates like flies and nematodes – differ significantly from the human eye in key anatomical and physiological features, especially in the cornea, lens, and retina9-12. The macula structure, a cone-rich region of the retina that sharpens central vision and colour perception, is also largely absent in many animal species, other than higher-order primates11. These differences restrain their functionality in modelling human eye diseases. Furthermore, many non-clinical disease models, such as those for laser‐induced choroidal neovascularization (CNV) in age-related macular degeneration (AMD), rely on artificially induced damage to mimic disease symptoms, rather than replicating the natural disease process. Such discrepancies in disease modelling and drug response can introduce high variability in preclinical studies and ultimately jeopardise translational success.
Alternative Testing Methods for Retinal Therapeutics
It is said that no model is perfect, but some provide valuable insights. When conducting preclinical studies, the critical questions become: is a model system useful? What are its limitations? How can we improve it?
In Vitro Retinal Models
Since the discovery of Yamanaka factors in 2006, which enabled the reprogramming of adult cells to induced pluripotent stem cells (iPSCs), human iPSCs have been widely used to generate various cell types and tissues for disease modelling and drug testing. hiPSC-derived retinal pigment epithelial (RPE) cells have become a standard tool for studying eye diseases in patients with diverse genetic backgrounds and conducting large scale drug screening. However, it remains controversial whether the reprogramming process erases epigenetic changes associated with cellular aging, potentially limiting the use of iPSCs to model aging-related diseases.
Apart from simpler 2D adherent cultures, in vitro retinal models are continuously progressing to self-organising micro-physiological systems, such as:
Human retinal organoids derived from hiPSC are 3D cellular aggregates that differentiate and self-organise to closely mimic the human retina13.
Co-culture systems mimicking blood-retinal barrier14 and retinal-immune cell interactions15.
Microfluidic eye-on-chip devices16 designed to resemble specialized tissue architecture.
Ex vivo culture of post-mortem human retinal tissue17 which preserves complex neuronal connections and retinal pathologies.
Advanced Computational Approaches
With the massive growth of multi-omics technologies, such as single-cell transcriptomics, metabolomics, proteomics, experimental and clinical datasets have become increasingly complex and high-dimensional. Computational approaches are now essential for aggregating and analysing these data to extract meaningful biological insights. This is especially important in highly heterogenous cell cultures differentiated from stem cells, such as hiPSC-derived RPE cells, where single-cell resolution is needed to characterise diverse cell states. By modelling the interactions between genes, proteins, and other biomolecules, these digital tools enable a more comprehensive understanding of retinal biology and disease mechanisms.
While most of these tools are not yet available for retinal therapeutic area, such applications have largely been developed for other therapeutic areas like oncology. AI and ML applications include:
Predicting drug efficacy and toxicity
Modelling ocular pharmacokinetics and drug metabolism
Anticipating off-target effects using structural biology data
Designing new drug candidates with improved safety and performance profiles
These advanced computational approaches will very soon be developed for retinal therapeutics and will complement cell-based models by offering powerful predictive tools that support the development of safer, more effective retinal therapies.
More will be discussed in our next editorial piece on the use of AI in accelerating drug discovery and development for retinal diseases.
Impact Analysis: What This Means for Retinal Therapeutics
The integration of cell-based platforms and advanced AI approaches could substantially reduce the reliance on animal models in retinal drug development. By providing more accurate representations for human retinal diseases, these technologies can improve the translation of results of preclinical testing into clinical outcomes, ultimately reducing late-stage trial failures and enabling more precise dosing and safety profiles.
The benefits extend beyond predictive accuracy:
Faster development timelines: Scalable non-animal systems can shorten preclinical phases and accelerate drug discovery and proof-of-concept studies.
Cost efficiency and lower barrier of entry for new retinal therapeutics companies: Eliminating the need for costly animal studies can reduce overall development costs and lower barriers for smaller biotech companies to enter the field.
Ethical and social acceptance: The alignment with growing public demand for humane research methods and sustainable research practices helps to enhance the social acceptance of retinal research outcomes.
Challenges and Strategic Considerations
One of the primary concerns lies in the current limitations of non-animal models, especially human cell-based systems. While these systems endeavour to mimic complex human physiology, validation are still undergoing to reliably predict long-term effects, toxicity, and therapeutic efficacy that can accurately translate to clinical outcomes18,19. For these models to be accepted in regulatory submissions, they must meet rigorous criteria, including reproducibility, sensitivity, specificity, and relevance to human biology, and through deep understanding on the transferability of each test20. Until these standards are universally met and accepted, regulatory agencies may hesitate to rely solely on such systems to make robust safety decisions.
The lack of harmonized international guidelines also adds complexity for global pharmaceutical development. Coordinated efforts and parallel initiatives with other regulatory agencies worldwide—such as the European Medicines Agency (EMA), China’s National Medical Products Administration (NMPA), Japan’s Pharmaceuticals and Medical Devices Agency (PMDA), and World Health Organization (WHO)—will be critical to ensure that new testing models are accepted across borders20. Singapore, Hong Kong and Taiwan could serve as crucial bridges between Western and Asian regulatory approaches as regional testing hubs, leveraging their established biotech ecosystems to develop and validate alternative testing methods.
Academic and clinical research institutions have a major role to play in this transformation by generating the foundational data needed to validate new testing platforms. Cross-disciplinary partnerships between academia, regulatory bodies, and industry will be essential to bridge knowledge gaps and accelerate model refinement.
Many pharmaceutical companies, including giants like Genentech-Roche, AstraZeneca, Eli Lilly and Pfizer, are already investing in modern AI-powered platforms as part of their long-term R&D strategies. A recent industry survey reported nearly half of pharmaceutical and biotech companies employ AI and digital tools for clinical research21. Early adopters could capture significant market share. Meanwhile, regulatory consultants and contract research organizations (CROs) are beginning to recalibrate their services and frameworks to support clients navigating this evolving regulatory landscape.
The Next Decade of Retinal Drug Development
The FDA Modernization Act 2.0 is poised to reshape R&D strategies in retinal therapeutics. As emerging technologies, like hiPSC-derived retinal organoids, organ-on-chip systems, and AI-driven modelling, gain traction, the potential to reduce reliance on animal models becomes increasingly feasible. While it may take several years before the first retinal drugs are approved under these new paradigms, the long-term impact could be profound: more predictive models, faster development cycles, and greater accessibility to novel treatments.
This transition calls for sustained efforts and transparent evaluation standards in validating alternative methods and maintaining scientific rigor. With continued collaboration across academia, industry, and regulators, the field is well-positioned to deliver safer, more effective retinal therapies to patients, sooner and more precisely than before.
References
1. Freedman LP, Cockburn IM, Simcoe TS. The Economics of Reproducibility in Preclinical Research [published correction appears in PLoS Biol. 2018 Apr 10;16(4):e1002626. doi: 10.1371/journal.pbio.1002626.]. PLoS Biol. 2015;13(6):e1002165. doi:10.1371/journal.pbio.1002165
2. Hay M, Thomas DW, Craighead JL, Economides C, Rosenthal J. Clinical development success rates for investigational drugs. Nat Biotechnol. 2014;32(1):40-51. doi:10.1038/nbt.2786
3. Dowden H, Munro J. Trends in clinical success rates and therapeutic focus. Nat Rev Drug Discov. 2019;18(7):495-496. doi:10.1038/d41573-019-00074-z
4. Hristodorov D, Lohoff T, Luneborg N, Mulder GJ, Clark SJ. Investing in vision: Innovation in retinal therapeutics and the influence on venture capital investment. Prog Retin Eye Res. 2024;99:101243. doi:10.1016/j.preteyeres.2024.101243
5. Zushin PH, Mukherjee S, Wu JC. FDA Modernization Act 2.0: transitioning beyond animal models with human cells, organoids, and AI/ML-based approaches. J Clin Invest. 2023;133(21):e175824. doi:10.1172/JCI175824
6. Library of Congress. (29 Sep 2022) S.5002 - FDA Modernization Act 2.0. 117th Congress (2021-2022). Retrieved 25 June 2025, from: https://www.congress.gov/bill/117th-congress/senate-bill/5002#:~:text=This%20bill%20authorizes%20the%20use,and%20effectiveness%20of%20a%20drug.
7. National Institutes of Health Office of Communications and Public Liaison. (29 April 2025) NIH to prioritize human-based research technologies. NIH News Releases. Retrieved 25 June 2025, from: https://www.nih.gov/news-events/news-releases/nih-prioritize-human-based-research-technologies
8. U.S. Food & Drug Administration. (10 April 2025). FDA Announces Plan to Phase Out Animal Testing Requirement for Monoclonal Antibodies and Other Drugs. FDA Press Announcements. Retrieved 25 June 2025, from: https://www.fda.gov/news-events/press-announcements/fda-announces-plan-phase-out-animal-testing-requirement-monoclonal-antibodies-and-other-drugs
9. Loiseau A, Raîche-Marcoux G, Maranda C, Bertrand N, Boisselier E. Animal Models in Eye Research: Focus on Corneal Pathologies. Int J Mol Sci. 2023;24(23):16661. doi:10.3390/ijms242316661
10. Horita S, Watanabe M, Katagiri M, et al. Species differences in ocular pharmacokinetics and pharmacological activities of regorafenib and pazopanib eye-drops among rats, rabbits and monkeys. Pharmacol Res Perspect. 2019;7(6):e00545. doi:10.1002/prp2.545
11. Buschbeck EK. Anatomical and functional diversity of animal eyes. In: Tsonis PA, ed. Animal Models in Eye Research. Academic Press; 2008:1-5. doi:10.1016/B978-0-12-374169-1.00001-1
12. Vezina M. Comparative Ocular Anatomy in Commonly Used Laboratory Animals. In: Weir A, Collins M, eds. Assessing Ocular Toxicology in Laboratory Animals. Humana Press, Totowa, NJ.; 2012:1-22. doi:10.1007/978-1-62703-164-6_1.
13. O'Hara-Wright M, Gonzalez-Cordero A. Retinal organoids: a window into human retinal development. Development. 2020;147(24):dev189746. doi:10.1242/dev.189746
14. Dujardin C, Habeler W, Monville C, Letourneur D, Simon-Yarza T. Advances in the engineering of the outer blood-retina barrier: From in-vitro modelling to cellular therapy. Bioact Mater. 2023;31:151-177. doi:10.1016/j.bioactmat.2023.08.003
15. Liu Y, Gao L, Chen W, Yan Y, Ye Z, Li Z. "Armed in-vitro retina"-generating microglial retinal organoids, where are we now?. Front Cell Dev Biol. 2025;13:1574283. doi:10.3389/fcell.2025.1574283
16. Maurissen TL, Spielmann AJ, Schellenberg G, et al. Modeling early pathophysiological phenotypes of diabetic retinopathy in a human inner blood-retinal barrier-on-a-chip. Nat Commun. 2024;15(1):1372. doi:10.1038/s41467-024-45456-z
17. Murali A, Ramlogan-Steel CA, Andrzejewski S, Steel JC, Layton CJ. Retinal explant culture: A platform to investigate human neuro-retina. Clin Exp Ophthalmol. 2019;47(2):274-285. doi:10.1111/ceo.13434
18. Lutolf MP, Radisic M, Beekman J, et al. In vitro human cell-based models: What can they do and what are their limitations? Cell. 2024;187(17):4439-4443. doi:10.1016/j.cell.2024.07.042
19. Han JJ. FDA Modernization Act 2.0 allows for alternatives to animal testing. Artif Organs. 2023;47(3):449-450. doi:10.1111/aor.14503
20. Courtot L, Fritsche E, Hobi N, et al. Panel discussions on the global regulatory acceptance and harmonisation of non-animal NAMs. NAM Journal. 2025;1:100027. doi:10.1016/j.namjnl.2025.100027
21. Thomas E. (6 December 2024) Half of pharma and biotech companies using AI and big data. Clinical Trials Arena. Retrieved 27 June 2025, from: https://www.clinicaltrialsarena.com/news/half-of-pharma-and-biotech-companies-using-ai-and-big-data/?cf-view&cf-closed