Rethinking Drug Safety
Each year, thousands of Australians are hospitalised by the very medicines meant to help them—a costly and largely preventable crisis.1 At the same time, many diseases and conditions still lack effective treatments.
This raises a critical question: How are medicines tested and approved—and why is the system failing on both safety and innovation?
How are drugs developed?
Australia’s Therapeutic Goods Administration (TGA) follows a drug development process similar to regulators worldwide:
- Discovery and development: Identify promising drug candidates.
- Preclinical testing: Typically involves animals to assess toxicity, basic safety and how the drug behaves in the body.
- Clinical trials (Phases I–III): Human testing in increasing scale to evaluate safety, effectiveness and optimal use.
- Regulatory review: Drug is submitted for approval.
- Post-market surveillance (Phase IV): Ongoing monitoring of safety in the wider population.
Despite scientific progress, this process is strikingly inefficient—taking 10 to 15 years and over US $2 billion to bring a single drug to market.2 More than a third of these costs occur before clinical trials, largely due to animal studies,3 which can be 1.5 to 30 times more expensive than modern lab-based methods.4
The role of animals in drug development
Animal testing became the norm after a 1937 disaster in the US, when a toxic antibiotic killed more than 100 people—prompting lawmakers to mandate animal safety checks, even though species differences weren’t well understood then (and remain a key problem today).
The numbers are staggering: a single chemical toxicity screen can involve around 5,000 animals;5 long-term cancer studies may use hundreds of rodents;6 and a typical monoclonal antibody program uses 144 non-human primates, with each animal costing up to US $50,000.7
In practice, the process ramps up dramatically: early screens unleash thousands of mice and rats, then escalate to dozens—or even hundreds—of dogs and non-human primates in later stages to trace a drug’s journey through the body, pinpoint organ damage and stress-test a vaccine’s protective power.
To cut costs and sidestep stricter welfare rules, many Australian labs outsource this work overseas—where oversight can be weakest. Worse still, much of this effort underpins ‘me-too’ drugs—minor tweaks on existing medicines that offer little new benefit but perpetuate unnecessary animal suffering.
Are animal models reliable?
Despite decades of reliance, animal-based research continues to produce poor outcomes. Around 92% of drugs that pass preclinical animal testing fail in human trials, often due to unexpected side effects or lack of effectiveness.8 Even after approval, one in three new drugs later face serious safety issues, including black box warnings, withdrawals or new restrictions.9
One of the key aims of preclinical testing is to detect toxicity—whether a drug causes harmful effects in the body. But animal models often fail to predict human responses. This can mislead researchers in two critical ways:
| Type of Error | Example | What Happened |
|---|---|---|
| Unsafe drugs that looked safe | Vioxx | Caused tens of thousands of heart attacks—undetected in animal trials. |
| Thalidomide | Linked to ~30,000 birth defects; showed no issues in over a dozen animal species. | |
| TGN1412 | Safe in animals at high doses, but caused life-threatening reactions in humans. | |
| Safe drugs that looked dangerous | Penicillin | Lethal to guinea pigs—nearly missed as a life-saving antibiotic. |
| Paracetamol | Toxic to dogs and cats, but safe and widely used in humans. | |
| Aspirin | Causes fetal harm in rats and rhesus monkeys, but commonly used in pregnancy. |
While animal research has played a role in past medical advances, its reliability for predicting human responses is increasingly in question—especially in an era of precision medicine, where treatments are tailored to individual biology.
Human-specific methods are the future
Around the world, scientists are embracing cutting-edge, human-relevant technologies:
- Organ-on-chip system
- 3D human tissue models
- AI-driven simulations
These tools are often faster, cheaper and more accurate than animal tests—and they’re already improving outcomes. Recent breakthroughs in cystic fibrosis, for example, have been made possible using patient-derived human tissue models.10
Regulators are starting to adapt. In the US, new legislation and funding programs support non-animal methods. The FDA has endorsed these technologies, signalling a global shift toward more human-relevant science.
It is a win-win for public health and ethics.’
Australia must catch up
Despite progress overseas, Australia still defaults to animal testing—with no formal plan to modernise. AFSA is a member of the International Council on Animal Protection in Pharmaceutical Programs (ICAPPP), working to harmonise global guidelines and reduce reliance on animals. We’re calling on the TGA to formally accept and promote non-animal methods in drug assessment.
Take action: Better tools for better medicines 📣
Currently, millions in taxpayer funding go to outdated approaches to medical research, while there is no dedicated funding stream for developing or validating modern, animal-free methods. The Australian Government must invest in human-specific research methods and provide a clear pathway for their regulatory use.
Here’s how you can help:
📧 Email your federal MP using our template to ask for funding for human-relevant drug testing.
🔬 Researchers, sign AFSA’s business case for investing in non-animal research methods.
Learn more
- Lim, R., Ellett, L. M. K., Semple, S. & Roughead, E. E. (2022). The extent of medication-related hospital admissions in Australia: A review from 1988 to 2021. Drug Saf, 45(3), 249-257. https://doi.org/10.1007/s40264-021-01144-1
- Lim, S. (2023). The process and costs of drug development (2022). FTLO Science. Accessed 12 May 2025. Available at: https://ftloscience.com/process-costs-drug-development/
- DiMasi, J. A., Grabowski H. G. & Hansen R. W. (2016). Innovation in the pharmaceutical industry: New estimates of R&D costs. J. Health Econ, 47, 20–33. https://doi.org/10.1016/j.jhealeco.2016.01.012
- Norman, G. A. V. (2019). Limitations of animal studies for predicting toxicity in clinical trials: Is it time to rethink our current approach? JACC Basic Transl Sci, 4, 845 https://doi.org/10.1016/j.jacbts.2019.10.008
- Badyal, D. K., & Desai, C. (2014). Animal use in pharmacology education and research: the changing scenario. Indian journal of pharmacology, 46(3), 257–265. https://doi.org/10.4103/0253-7613.132153
- Manuppelo J., Slankster-Schmierer E., Baker E. & Sullivan K. (2023). Animal use and opportunities for reduction in carcinogenicity studies supporting approved new drug applications in the U.S., 2015-2019. Regulatory Toxicology and Pharmacology, 137, 105289. https://doi.org/10.1016/j.yrtph.2022.105289
- FDA. (2025). Roadmap to reducing animal testing in preclinical safety studies. https://www.fda.gov/media/186092/download?attachment
- Marshall, L. J., Bailey, J., Cassotta, M., Herrmann, K. & Pistollato, F. (2023). Poor translatability of biomedical research using animals – A narrative review. ATLA, 51(2), 102–135. https://doi.org/10.1177/02611929231157756
- Downing, N. S., Shah, N. D., Aminawung, J. A., Pease, A. M., Zeitoun, J. D., Krumholz, H. M. & Ross, J. S. (2017). Postmarket safety events among novel therapeutics approved by the US food and drug administration between 2001 and 2010. JAMA, 317(18), 1854–1863. https://doi.org/10.1001/jama.2017.5150
- Marshall L.J. & Conlee K.M. (2024). The case of the missing mouse—developing cystic fibrosis drugs without using animals. Front. Drug Discov. 4:1347246. https://doi.org/10.3389/fddsv.2024.1347246
Support our campaigns
Share us on social
Buy cruelty-free
Share via Social Media
Stay informed
Sign up to receive all the latest news and updates with our e-newsletter.Prefer to be notified via SMS? Click here to opt-in for SMS alerts.
"*" indicates required fields