Understanding Drug-Drug Interactions: A Call for Human-Relevant Tools
In today’s world of complex therapeutic regimens and polypharmacy, the ability to accurately predict drug-drug interactions (DDIs) has become more critical than ever. Whether it’s a cancer patient taking multiple targeted therapies or an elderly individual managing chronic diseases with various prescriptions, the risk of unintended drug interactions affects millions of patients worldwide. Studies have shown that over 22% of hospital admissions related to adverse drug reactions are caused by drug-drug interactions, highlighting their significant contribution to preventable hospitalizations.
DDIs can significantly alter a drug’s efficacy or safety by impacting its absorption, metabolism, or clearance. These interactions can lead to therapeutic failure, serious adverse events, increased hospitalization rates, or even long-term harm to patient health. When undetected early, these interactions can also contribute to clinical trial challenges, regulatory delays, or post-approval safety issues, creating substantial consequences for both patients and pharmaceutical developers.
Why This Matters: The High Stakes of Overlooking DDIs
Drug-drug interactions (DDIs) are not rare, they are a widespread and persistent challenge across therapeutic areas, particularly in complex treatment regimens. Clinically, significant risks have been documented with antiviral therapies, including those used to treat hepatitis C, where co-administration with other medications has led to serious interaction concerns. For example, statins such as simvastatin (cholesterol-lowering agents) can increase the risk of muscle toxicity when used alongside certain antivirals, and antipsychotics like quetiapine can heighten sedation and toxicity risks. These risks extend beyond antivirals, affecting oncology, cardiovascular, and psychiatric therapies particularly in patients managing multiple prescriptions. Moreover, as cancer treatments and mental health therapies grow increasingly complex, the potential for clinically significant DDIs continues to rise. Each missed DDI represents more than a regulatory or financial setback; it reflects a critical gap in patient protection. Yet despite growing awareness, many interactions are still detected too late, often only after adverse events occur. Addressing this challenge requires preclinical in vitro systems that can monitor how drug behavior changes over time when given in combination, capturing delayed effects, metabolic shifts, and the downstream consequences that conventional in vitro systems often miss.
Where Traditional Models Fall Short
Despite decades of development, the in vitro systems most widely used for DDI prediction offer limited mechanistic insight. Suspension hepatocytes, static cultures, and immortalized liver cell lines often fail to sustain hepatic function beyond a few days. This rapid loss of metabolic function limits static cultures to only short-term induction studies, making it difficult to evaluate chronic or delayed DDI effects. Attempts to extend hepatocyte function using stromal co-culture introduce non-human cells, which compromise human specificity and may confound clearance predictions.
Animal models, meanwhile, offer only partial answers. Species differences in metabolism, transporter expression, and enzyme induction mean that preclinical findings often fail to translate reliably to human outcomes.
As a result, drug developers are often left with a critical blind spot: while static systems can measure short-term effects, they present significant challenges in capturing how drug interactions evolve dynamically over time or across patient populations. Bridging this gap requires more than extending the lifespan of cultures, it demands a fundamental shift in how we model drug interactions over time.
Rethinking DDI Prediction Through Mechanistic Insight
Accurately capturing DDIs, especially those with delayed, reversible, or multi-pathway effects requires more than detecting short-term changes in enzyme activity. As drug combinations become more complex, the ability to understand not just if an interaction occurs but when and how becomes increasingly important. This calls for systems that sustain functional expression of both enzymes and transporters, support both single-dose and repeat dosing regimens, and allow researchers to track how interactions emerge and resolve over time. Without these core capabilities, key interaction risks, particularly those involving metabolism, transport, or prolonged exposure often remain hidden until late-stage development.
To truly reflect how drugs behave in the body, we need systems that allow for chronic dosing, where researchers can track changes in drug behavior over time. This includes the ability to sample over time - before, during, and after the introduction of another drug, so that kinetics and tissue response can be evaluated with greater resolution.
There is also a growing need to study interactions involving pathways beyond just CYP enzymes. In many cases, clinically significant DDIs arise from non-CYP enzymes such as UGTs and transporters like OATP1B1 or P-gp. These components are often unstable or absent in conventional in vitro systems, but they play a major role in how drugs are absorbed and cleared. Without the right tools, developers may miss important interaction risks, especially for drugs that rely heavily on active transport or undergo metabolism through multiple pathways. These limitations leave major blind spots. A drug might seem safe after a single dose or shorter duration incubation, but could have delayed or lasting effects that only show up with longer studies. What’s needed is a system that can simulate chronic dosing, handle multiple exposure scenarios, and allow researchers to observe how induction develops and resolves over time.
Another underappreciated but critical dynamic is the reversibility of enzyme induction. In clinical settings, enzyme activity doesn’t remain elevated forever, it recovers once the inducer is removed. Modeling this return to baseline, often referred to as de-induction kinetics, is essential to understand how long a DDI might persist after treatment ends. But this type of time-course assessment simply isn’t possible in most short-lived cultures.
Emerging tools that enable such longitudinal studies, especially those that support chronic dosing and accommodate transporter and non-CYP mechanisms, are reshaping how we approach DDI risk early in development.
How Javelin’s Platform Enables This Shift
To meet these challenges, Javelin has developed a human-relevant liver tissue platform designed for long-term DDI studies. Our system supports multi-week liver function in a human-only culture and maintains robust CYP activity across several isoforms, along with measurable transporter activity. The platform’s stable configuration enables extended culture periods without requiring frequent medium changes, allowing multiple time points to be sampled from the same chip. This capability supports tracking of metabolite accumulation over time, which is critical for identifying delayed, secondary, or metabolite-mediated DDI risks—as well as complex interactions involving enzyme-transporter interplay or emerging modalities like antisense oligonucleotides (ASOs) and siRNA therapies that challenge traditional prediction systems.
Researchers can track the induction of enzymes like CYP3A4, CYP2C9, and CYP2C19 across different dosing strategies, and then continue observing activity even after the inducer is withdrawn. This ability to follow both onset and recovery of enzyme changes gives a more complete picture of risk, particularly for drugs that are dosed in cycles or intermittently.
In our recent study, we demonstrated that the platform captures clinically relevant changes in victim drug clearance following enzyme induction. For example, rifampicin treatment led to a two-fold increase in midazolam clearance and a 2.6-fold increase for alprazolam, mirroring clinical trends for both high and low-clearance compounds. These results highlight the platform’s ability to support long-term induction studies and generate clinically aligned DDI outcomes across diverse metabolic profiles. Importantly, longitudinal metabolite profiling, such as increased 1-OH midazolam formation, further supports the system’s utility in identifying metabolite-driven DDI risks and potential toxicities early in development.
By maintaining robust enzyme and transporter activity over extended periods, our platform enables detection of transporter-based DDIs and complex metabolic interactions that are often missed in traditional short-term models. Integration with gut–liver configurations further supports the evaluation of first-pass effects in a physiologically relevant context.
As therapeutic regimens become increasingly complex, the need for human-relevant, longitudinal evaluation of DDI risks has never been greater. This is essential not only for advancing safer therapies, but for upholding the trust patients place in modern medicine. Javelin’s liver platform meets this challenge by enabling researchers and developers to generate mechanistic insight, design more predictive development strategies, and make more confident decisions earlier in the pipeline. Together, by investing in better predictive models, we can deliver safer, smarter treatments to the patients who need them most.