How real-world safety data could unlock new opportunities for drug repurposing
Drug development has never been more scientifically sophisticated – or more challenging. Despite advances in genomics, AI, and precision medicine, bringing a new therapy from discovery to market remains a lengthy, costly, and high-risk endeavor. Estimates suggest that developing a novel medicine can take more than a decade and cost billions of dollars, with the majority of candidates failing before reaching patients.
Against this backdrop, drug repurposing has emerged as an increasingly attractive strategy. By identifying new therapeutic uses for existing medicines, organizations can potentially reduce development timelines, leverage established safety profiles, and accelerate patient access to innovative treatments. Yet one critical challenge remains: where do promising repurposing opportunities come from?
Traditionally, repurposing candidates have been identified through laboratory research, clinical observation, or advances in disease biology. However, another valuable source of insight may already exist within datasets generated every day by the pharmaceutical industry itself.
Pharmacovigilance (PV) databases contain millions of real-world reports submitted by healthcare professionals, patients, and manufacturers worldwide. These systems are typically used to identify safety risks, but they may also contain clues about unexpected therapeutic benefits. As the industry increasingly embraces real-world evidence and advanced analytics, an important question is emerging: can pharmacovigilance data contribute not only to patient safety, but also to therapeutic innovation?
Our recent research exploring antidiabetic drugs as potential candidates for inflammatory bowel disease (IBD) provides a compelling example of how pharmacovigilance may help answer that question.
What Is Inverse Signal Detection?
Traditional pharmacovigilance relies heavily on disproportionality analysis. Metrics such as the Reporting Odds Ratio (ROR), PRR (Proportional Reporting Ratio) and EBGM (Empirical Bayes Geometric Mean) are used to identify adverse events reported more frequently than expected for a particular drug. These positive signals help detect potential safety concerns and form the backbone of modern signal management practices.
Inverse signal detection applies the same analytical principles in reverse.
Rather than searching for adverse events reported more frequently than expected, inverse signal detection seeks drug-event combinations reported less frequently than anticipated. When a statistically significant inverse association is observed, it may suggest a protective effect, beneficial biological activity, or an opportunity for therapeutic repositioning.
Importantly, inverse signals do not demonstrate efficacy or causality. Instead, they function as hypothesis-generating observations. They help researchers identify patterns that warrant further investigation through mechanistic studies, epidemiological analyses, or clinical research.
The value of this approach lies in its efficiency. Large spontaneous reporting databases such as the FDA Adverse Event Reporting System (FAERS) already contain vast amounts of real-world information across thousands of products and therapeutic areas. Hidden within these datasets may be relationships that would otherwise remain undetected using conventional drug discovery approaches.
For pharmaceutical organizations facing increasing pressure to improve R&D productivity, inverse signal detection offers a potentially powerful way to generate new ideas from existing data assets.
The Study at a Glance
This study investigated whether inverse signal detection could identify potential drug repurposing candidates for inflammatory bowel disease using data from FAERS.[1]
IBD, which includes Crohn’s disease and ulcerative colitis, remains a significant therapeutic challenge. While biologics and targeted therapies have improved outcomes for many patients, substantial unmet needs persist. Loss of response, incomplete remission, safety concerns, and treatment costs continue to drive demand for new therapeutic approaches.
Using FAERS data spanning nearly two decades, we conducted a systematic screening process designed to identify medications demonstrating statistically significant inverse associations with IBD-related adverse event reporting.
The analysis began with thousands of potential drug-event combinations. Through successive filtering based on disproportionality metrics, confidence intervals, and multiple-testing corrections, 73 candidate drugs were ultimately identified.
Among these candidates, antidiabetic medications emerged as one of the most intriguing therapeutic categories.
Ten antidiabetic agents demonstrated significant inverse associations and were prioritized for further evaluation. The findings included representatives from several drug classes, suggesting that the observed pattern was unlikely to be attributable to a single product alone.
Perhaps most importantly, the results aligned with an expanding body of scientific literature highlighting connections between metabolic regulation, inflammation, immune function, and gut health.
Key Findings: Why Antidiabetic Drugs Matter in IBD
Historically, diabetes and inflammatory bowel disease have been treated as distinct disease areas. However, advances in immunology, metabolism research, and microbiome science have revealed increasingly complex interactions between these conditions.
Chronic inflammation plays a central role in both metabolic disorders and immune-mediated diseases. Similarly, gut microbiota composition, intestinal barrier function, cytokine signaling, and immune regulation influence disease progression across multiple therapeutic areas.
Many modern antidiabetic therapies exhibit biological effects extending beyond glucose control.
GLP-1 receptor agonists, for example, have attracted substantial attention for their anti-inflammatory and immunomodulatory properties. Experimental studies suggest these agents may reduce inflammatory cytokine production, improve intestinal barrier integrity, and influence immune responses relevant to IBD pathogenesis.
Other antidiabetic drug classes have also demonstrated mechanisms that may support gastrointestinal health, including effects on oxidative stress, metabolic signaling pathways, and microbiome composition.
While our findings do not establish therapeutic efficacy, they reinforce growing evidence that metabolic pathways may represent an important and underexplored target in inflammatory bowel disease.
This broader perspective may ultimately prove more valuable than any individual drug candidate. The identification of multiple antidiabetic agents points toward a therapeutic hypothesis rather than a single product opportunity. Such insights can help researchers prioritize future studies and identify novel biological pathways worthy of exploration.
Implications for Pharmacovigilance
The most significant implication of this research may not be the specific drugs identified, but rather what the findings suggest about the evolving role of pharmacovigilance.
For decades, PV has primarily been viewed as a post-marketing safety discipline focused on risk identification, assessment, and mitigation. That mission remains essential. However, the rapid growth of real-world data is expanding the potential contribution of pharmacovigilance across the product lifecycle.
Inverse signal detection illustrates how safety databases can also generate innovation-focused insights.
By leveraging existing pharmacovigilance datasets, organizations may be able to identify promising therapeutic hypotheses at relatively low cost. This approach complements traditional discovery methods and supports a more data-driven model of pharmaceutical innovation.
The concept aligns closely with broader industry trends. Real-world evidence is increasingly informing regulatory decisions, clinical development strategies, and health technology assessments. Artificial intelligence and advanced analytics are transforming how organizations extract value from complex datasets. Within this environment, pharmacovigilance databases represent a largely untapped resource for therapeutic discovery.
The findings also highlight the growing convergence of pharmacovigilance, epidemiology, translational science, and data analytics. Future PV professionals may play a broader role than ever before-not only safeguarding patients but also contributing to scientific innovation and strategic decision-making.
Of course, important limitations remain. Spontaneous reporting systems are subject to reporting biases, confounding factors, underreporting, and variable data quality. As a result, inverse signals should never be interpreted as proof of efficacy.
Nevertheless, when integrated with biological plausibility, observational research, and clinical investigation, inverse signal detection can serve as a valuable tool for prioritizing opportunities and generating novel therapeutic hypotheses.
Conclusion
Drug repurposing continues to offer one of the most promising pathways for accelerating therapeutic innovation while reducing development risk. Our study demonstrates how pharmacovigilance databases can contribute to this process by identifying unexpected relationships between established therapies and unmet clinical needs.
The emergence of antidiabetic drugs as potential candidates for inflammatory bowel disease highlights the value of looking beyond traditional safety applications of pharmacovigilance data. More broadly, it demonstrates how inverse signal detection can transform large-scale real-world datasets into actionable scientific insights.
Further validation through laboratory, epidemiological, and clinical studies remains essential. However, the broader message is clear: pharmacovigilance may have a much larger role to play in the future of drug discovery than previously recognized.
As pharmaceutical organizations continue to seek faster, smarter, and more cost-effective approaches to innovation, the next therapeutic breakthrough may not come solely from new data-but from finding new ways to interpret the data we already have.
Author Bio
Katarina Đogatović, MD, is an Associate Medical Director at PrimeVigilance Ltd. with more than seven years of experience in drug safety and risk management and a PhD candidate in Clinical Pharmacy and Pharmacokinetics. Her research focuses on leveraging pharmacovigilance and real-world data for drug repurposing, particularly through inverse signal detection methodologies and their application in inflammatory bowel disease.
Prioritizing Antidiabetic Drugs for Inflammatory Bowel Disease Through Inverse Signal Detection: A FAERS Pharmacovigilance Study
[1] Đogatović K, Vučićević K, Marković S, Kovačević M, Ćulafić M, Miljković B, Vezmar Kovačević S. Prioritizing Antidiabetic Drugs for Inflammatory Bowel Disease Through Inverse Signal Detection: A FAERS Pharmacovigilance Study. Journal of Clinical Medicine. 2026; 15(12):4672. https://doi.org/10.3390/jcm15124672
