In 1968, Schwab et al reported an unexpected result in their trial of amantadine, a tricyclic amine compound initially developed for the treatment of Asian influenza: one of their patients exhibited marked dyskinesia improvement related to their Parkinson’s Disease. Once the course was finished, however, the symptoms returned.
This observation contributed to a massive uptick in pre-clinical and clinical trial interest during the ensuing decades, and now amantadine is not only the sole treatment with proven efficacy for levodopa-induced dyskinesias, but also used off-label for Huntington’s Disease, tardive dyskinesia, and other movement disorders.
Ultimately, Schwab’s team uncovered an unexpected boon to patients: an effective treatment that might never have been developed.
This is serendipitous drug discovery, powered by human observation and intelligence. While artificial intelligence (AI) may be surpassing humans in specific, data-intensive tasks, humans still show greater aptitude in terms of creativity, deductive reasoning and ethical judgment. With the use of modern approaches and tools, such as real-world data analytics, and yes, AI-powered solutions, we can intentionally expand our observational power, bringing an analytical mind to serendipitous drug discovery.
Comprehensive Statistics
Although it’s been over a decade since any significant comprehensive research has gone into serendipitous drug development, we have some key core statistics to consider. As of 2012:
- 5.8% of marketed drugs involved serendipitous events
- 24.1% include chemical derivatives
- Breakthroughs in labs vs. clinical discovery is 2.2% to 3.7%
- CNS drugs, on average, have had the highest benefit
- Cancer drugs also see over 35% connection to serendipity
More serendipity
- Penicillin: Alexander Fleming observed mold-killing bacteria in a petri dish after returning to his untidy lab after a two-week holiday
- Sildenafil (Viagra): Originally developed to treat angina, nurses observed a benefit for patients with erectile dysfunction
- Thorazine: Initially explored for surgical shock in the 1950s, French psychiatrists found that it calmed psychotic patients, marking the dawn of psychopharmacology
All told, it’s estimated that 1 in 4 drugs can trace their origin to serendipity. And discoveries often happen across therapeutic classes.
The economic impact tells its own story. Viagra alone generated $2.9 billion in 2023. Perhaps even more compelling is the recent story of GLP-1 receptor agonists. Originally developed for type 2 diabetes, these drugs showed unexpected weight loss benefits that were initially considered side effects. Today, medications like semaglutide have revolutionized obesity treatment, with Wegovy and Ozempic becoming household names. The observation that diabetic patients were losing significant weight led to an entirely new therapeutic category—one that’s projected to become a $100 billion market by 2030. It’s a perfect example of how paying attention to “off-target” effects can unlock massive therapeutic and commercial potential.
Embracing Observation and Nurturing Serendipity
When Schwab et al observed dyskinesia improvement in Parkinson’s patients, they chose the path of curiosity and improved patient outcomes. It happened because their patients were being treated for comorbid conditions, and the care teams were paying close attention to off-target effects. This, in turn, helped build a system which allowed real-world observation to influence the research.
So what do these serendipitous discoveries have in common?
- Real-world environments where patients have comorbidities
- Researchers and clinicians who observed “off-target” effects and pursued them
- Organizations willing to pivot when new evidence emerges
Fewer than 10% of drug candidates succeed in clinical trials. That number has not changed much in decades, and complex diseases remain difficult to treat and expensive for the patient, the payer, and the institutions. We are now seeing narrowing inclusion/exclusion criteria in clinical trial design, with time-limited periods for observation, in the push to get drugs to market faster and with cleaner data. With focuses bent on primary endpoints, the incorporation of real-world observation becomes even more important.
Real-world data (RWD) provides opportunities to accelerate timelines and save money without compromising flexibility. Common examples include the use of RWD to inform clinical study design and to understand the economic burden of disease, as well as post-marketing opportunities to analyze drug adherence and persistence versus competitors.
Here’s the paradox: we have more sophisticated tools than ever before, yet we’re not seeing the breakthrough rate we might expect. A 2024 analysis in Nature Computational Science found that machine learning algorithms haven’t delivered the anticipated leap forward in drug discovery.
Long-Term Follow-Up and Real-World Data Integration
Traditional clinical trials often end too soon to capture emerging benefits, so the industry has had to adapt. Long-term follow-up (LTFU) studies, registries, and real-world observational studies designed for signal detection can extend the discovery window far beyond the initial controlled environment.
This is where modern data analytics become invaluable. It’s relatively common now to use RWD to identify off-label use of drugs for the purposes of label expansion. Healthcare utilization patterns could show fewer emergency room visits or specialist appointments. Pharmacy and medical claims and electronic health records can reveal real-world patterns in medication use. Going one step further, by tokenizing and linking trial participants to RWD sources like claims and EHRs, we can extend observation of clinical trial participants into the real world. It’s still observation—primed for serendipity, if you will—but with thoughtful design applied, the benefits are very promising.
Consider how such tools might capture the amantadine discovery today. Instead of chance observation leading to discovery, systematic data analysis would reveal a subpopulation of patients being treated for influenza who have also been diagnosed with Parkinson’s disease. An analysis of that subpopulation could reveal improvements in Parkinson’s symptoms. The signal would emerge in the dataset, enabling a targeted follow-up trial designed based on real evidence rather than anecdotal reports.
The Psychedelic Opportunity
Psychedelic research gives us a unique opportunity to build discovery-friendly systems from the ground up, for several compelling reasons. These compounds show promise across multiple indications—depression, anxiety, PTSD, addiction, OCD, eating disorders—making them perfect candidates for broad assessment. Patients consistently report holistic life improvements that extend beyond target symptoms. Benefits often emerge over time, sometimes months after treatment. Individual response varies significantly, suggesting multiple mechanisms at play.
And here’s the key advantage: psychedelic therapies already require comprehensive monitoring throughout the treatment journey and beyond. Rather than adding burden, discovery-friendly design simply channels existing observation into systematic data collection. So the burden is relatively minimal.
During acute treatment sessions, therapists and medical monitors are already present for safety. Adding structured documentation of unexpected psychological insights or emotional breakthroughs requires minimal additional effort. Integration sessions—the therapy sessions following psychedelic experiences—naturally involve patients describing what’s changed in their lives. It’s baked into the process. Systematically capturing all reported changes, not just those related to the primary indication, becomes a matter of good documentation, not adaptation.
Long-term registries tracking participants for 6-24 months or longer could assess improvements across life domains: not just depression scores, but also substance use patterns, relationship quality, sense of meaning and purpose, chronic pain levels, and physical health markers. Many psychedelic studies already include broad assessment batteries. The difference is intentionality when it comes to uncovering unexpected benefits.
Early signals are already emerging, often buried in secondary endpoint data or patient testimonials. Depression trials are showing anxiety improvement that rivals the primary outcome. PTSD treatment appears to affect problematic substance use. Addiction treatment shows depression lifting. Anecdotal reports describe chronic pain improvement, better sleep, enhanced creativity, improved relationships.
These shouldn’t remain anecdotes. With the right framework, they become hypotheses worth testing.
The business case for psychedelic developers is compelling. The same infrastructure that captures safety and efficacy data can capture discovery signals with marginal additional cost. The potential payoff? dramatically improving the probability of commercial success and patent life through indication expansion.
It’s also a powerful demonstration of patient-centered commitment—an argument that resonates with payers increasingly focused on holistic outcomes rather than symptom checklists. In a competitive landscape where multiple companies are pursuing similar compounds, discovery efficiency could provide decisive advantage.
The Patient as North Star: A Conclusion
There’s also an ethical imperative here that goes beyond methodology. Patients donate their time, accept risks, and often travel great distances to participate in clinical trials. They do this hoping to help themselves and future patients. We owe them comprehensive observation of their experience. Dismissing unexpected benefits because they weren’t in the protocol isn’t just scientifically wasteful—it’s disrespectful to the contribution participants make.
That nurse who noticed the Parkinson’s tremor improving in a patient being treated for flu wasn’t doing anything revolutionary. She was simply paying attention to the whole patient, not just the indication being treated. That is real patient centricity, not lip service.
Researchers and even doctors sometimes forget that patients don’t experience disease in neat categories. They experience their lives—all of it, all at once, in stages, during flare ups, or sometimes not at all. Depression doesn’t politely separate itself from chronic pain, or anxiety from insomnia, or addiction from trauma. Not every chronic illness poses problems 24/7. Many patients experience invisible illnesses and sustained psychiatric stress. When we design trials that only look at isolated symptoms, we’re not seeing patients. We’re seeing fragments conveniently partitioned to better reflect an already siloed healthcare system.
Patients should be our North Star, not the protocol. Protocols should serve patients, not the other way around. When a patient tells us something unexpected is happening, our response shouldn’t be “that’s not what we’re measuring.” It should be “tell me more.”
Serendipity, in this light, isn’t really about luck at all. It’s about listening to patients and validating their experiences. Even if they don’t immediately seem important.
In psychedelic medicine. the therapy IS the drug—the experience, the insights, the psychological shifts, the integration work. Which means the patient’s full experience is the data.
Perhaps most importantly, we acknowledge a fundamental truth that every clinician knows but the system often forgets: healing happens in patterns we don’t always predict, through mechanisms we don’t fully understand, in timeframes that don’t fit our protocols. It is unpredictable, but that doesn’t mean it’s chaos: it means, in many cases, it’s opportunity.
Bibliography
21st Century Cures Act. U.S. regulatory framework.
Ban, Thomas A. “The Role of Serendipity in Drug Discovery.” Dialogues in Clinical Neuroscience 8, no. 3 (2006): 335–344. https://doi.org/10.31887/DCNS.2006.8.3/tban.
Ban, Thomas A., David Healy, and Edward Shorter. “Serendipity in Anticancer Drug Discovery.” World Journal of Clinical Cases 1, no. 3 (2012): 91–96.
Clinical Trials Arena. “Most Significant Clinical Trials of 2024.” Accessed 2025. https://www.clinicaltrialsarena.com/features/most-significant-trials-2024/.
CNBC. “What’s Next for the Weight Loss Drug Market: Pills, Rivals, Insurance.” November 2, 2025. https://www.cnbc.com/2025/11/02/whats-next-for-the-weight-loss-drug-market.
“Dynamic Clinical Trial Success Rates for Drugs in the 21st Century.” Nature Communications (2025). https://doi.org/10.1038/s41467-025-64552-2.
Englebert. “Off-Label Drug Use Liability.” The Regulatory Review, June 13, 2024. https://www.theregreview.org/2024/06/13/englebert-off-label-drug-use-liability/.
Fleming, Alexander. Original penicillin discovery, 1929.
Global Markets Insights. “Erectile Dysfunction Drugs Market Size, Share & Growth Report 2025–2034.” 2024. https://www.gminsights.com/industry-analysis/erectile-dysfunction-drugs-market.
Goldman Sachs Research. “Why the Anti-Obesity Drug Market Could Grow to $100 Billion by 2030.” 2024. https://www.goldmansachs.com/insights/articles/anti-obesity-drug-market.
IQVIA Institute. “Global Trends in R&D 2024: Activity, Productivity, and Enablers.” 2024. https://www.iqvia.com/insights/the-iqvia-institute/reports-and-publications/reports/global-trends-in-r-and-d-2024.
J.P. Morgan Research. “The Increase in Appetite for Obesity Drugs.” 2024. https://www.jpmorgan.com/insights/global-research/current-events/obesity-drugs.
Market Data Forecast. “Minoxidil Market Size, Share & Growth Report, 2033.” 2024.
Nature Computational Science. Various articles on machine learning in drug discovery, 2024. https://www.nature.com/subjects/machine-learning/natcomputsci.
ResearchAndMarkets.com. “GLP-1 Agonists Market—Global Forecast to 2033.” Business Wire, 2025.
“The Role of AI in Drug Discovery: Challenges, Opportunities, and Strategies.” PMC (2023). PMCID: PMC10302890.
Schwab, Robert S., Arthur C. England, David C. Poskanzer, and Robert R. Young. “Amantadine in the Treatment of Parkinson’s Disease.” JAMA 208, no. 7 (1969): 1168–1170.
Spherical Insights. “Global Minoxidil Market Size, Share, and Forecasts to 2033.” 2024. https://www.sphericalinsights.com/reports/minoxidil-market.
U.S. Food and Drug Administration. “Compilation of CDER New Molecular Entity (NME) Drug and New Biologic Approvals.” 2024. https://www.fda.gov/drugs/drug-approvals-and-databases/compilation-cder-new-molecular-entity-nme-drug-and-new-biologic-approvals.
U.S. Food and Drug Administration. Guidance Documents on Real-World Evidence.
Viatris Inc. “Fourth-Quarter and Full-Year 2023 Financial Results.” Press release, February 28, 2024. https://newsroom.viatris.com/2024-02-28-Viatris-Reports-Fourth-Quarter-and-Full-Year-2023-Financial-Results.
Wang, Y., et al. “Limitations of Representation Learning in Small Molecule Property Prediction.” Nature Communications (2023). https://doi.org/10.1038/s41467-023-41967-3.
Wikipedia. “Amantadine.” Accessed 2025. https://en.wikipedia.org/wiki/Amantadine.
Within3. “6 Essential Pharmaceutical Industry Statistics to Know in 2024.” 2024. https://within3.com/blog/pharmaceutical-industry-statistics.
About UBC
United BioSource LLC (UBC) is the leading provider of evidence development solutions with expertise in uniting evidence and access. UBC helps biopharma mitigate risk, address product hurdles, and demonstrate safety, efficacy, and value under real-world conditions. UBC leads the market in providing integrated, comprehensive clinical, safety, and commercialization services and is uniquely positioned to seamlessly integrate best-in-class services throughout the lifecycle of a product.
About the Author
Judy Lytle, PhD, MBEE, PMP. Executive Director, Evidence Development Study Solutions
Judy Lytle serves as the Executive Director of Evidence Development Study Solutions for UBC. Dr. Lytle joined UBC in 2023, bringing more than 15 years of experience in life science and healthcare strategy development, implementation, and execution. With a background in medical affairs and real-world evidence, she brings together differentiated study design and evidence generation solutions for value demonstration. She also has oversight of epidemiology, patient and physician services, scientific/clinical strategy, and medical writing teams.
Dr. Lytle holds a PhD in Neuroscience from Georgetown University as well as a Master of Biotechnology Enterprise & Entrepreneurship (MBEE) from Johns Hopkins University. A fellow of the American Association for the Advancement of Science (AAAS), and certified Project Management Professional (PMP), her approach is systematic and grounded in science.
Natania Barron, Sr. Director, Marketing
Natania Barron is a life sciences marketing professional with over 20 years of experience. Her passion is storytelling at the intersection of data and narrative.
