As pharmaceutical research pushes towards previously “undruggable” targets, the challenge often boils down to one key issue: complex drugs with poor water solubility.
To overcome this obstacle, scientists typically turn to Amorphous Solid Dispersions (ASDs) to enhance solubility and bioavailability. However, these systems introduce their own complexity. One major hurdle is the so-called “limit of congruency” (LoC). The LoC marks the drug loading threshold beyond which drug release suddenly and dramatically declines – a point researchers often call “falling-off-the-cliff”. This tipping point is paramount, as crossing it transforms a promising formulation into one that no longer delivers the drug as intended.
Recognizing the critical need for a predictive solution, scientists from Purdue University, AbbVie Inc., and amofor GmbH teamed up with a clear goal: Develop a reliable, physics-based model to forecast at which drug load the LoC lies- and understand the mechanism resulting in this release clash. Their findings, published in Molecular Pharmaceutics, mark a major step forward in understanding and forecasting the LoC, transforming how researchers can optimize ASD formulations.
To understand the significance of this breakthrough, we spoke with Dr. Christian Lübbert from amofor to discuss the collaboration and the findings of this study.
Why is it important to be able to assess and predict drug release?
Christian: Solubility is the linchpin of effective drug therapy. A drug that does not dissolve properly cannot be absorbed, which reduces its bioavailability and limits its therapeutic effect. Knowing the release profile of an active pharmaceutical ingredient enables us to identify obstacles in development at an early stage and optimize formulation strategies. But with ASDs, the release profiles have been unpredictable so far.
Before hitting the LoC, an ASD allows the drug to form discrete nanodroplets. These dissolve rapidly and ensure robust release. Once past that limit, however, the drug becomes trapped in a continuous, hydrophobic layer. This sharply cuts off release. Historically, formulators had to rely on exhaustive trial-and-error experiments to identify these tipping points empirically. This process costs time, money, and resources. It also does not provide a mechanistic understanding of what is going on in the formulation – an approach unsuitable for the high demands of modern drug development.
A reliable in silico model for predicting the LoC can eliminate costly guesswork, speed up development, and increase the likelihood of finding stable and effective formulations from the start.
Which methods have you used to address this challenge?
Christian: To address the LoC, we combined extensive and meticulous experimental work with advanced physics-based modeling. Colleagues at Purdue University prepared ASDs with varying drug loadings and polymer concentrations – focusing on PVPVA64, a very common hydrophilic, typical ASD polymer. This polymer is known for its rapid release, followed by a sharp concentration drop. They exposed these formulations to carefully controlled humidity and aqueous conditions and tracked evolving phase compositions using NMR spectroscopy and fluorescence confocal microscopy. With the experimental data, they created a set of ternary phase diagrams that clearly show where the drug, polymer, and water remain miscible or separate.
In parallel, our team employed a PC-SAFT (Perturbed Chain Statistical Associating Fluid Theory) based modeling technique to calculate these diagrams computationally. By accounting for molecular interactions like hydrogen bonding and van der Waals forces, PC-SAFT provides a robust thermodynamic framework for these interactions.
What are the key findings of this study?
Christian: In this study, we forecasted exactly where the LoC in those PVPVA64-containing ASDs occurs and how this correlates with the phase diagram – essentially, the “sweet spot” where the API begins to precipitate. Our findings provide a clear, mechanistic link between drug-polymer interactions, water penetration, amorphous phase separation, and dissolution behavior.
For instance, we found that strong drug-polymer interactions shift the critical point and trigger phase inversion at lower drug loadings, which explains why some ASDs “fall off the cliff” at relatively modest API drug loads. This happens particularly in PVPVA64-based ASDs.
Importantly, confirmatory microscopy experiments revealed that formulations below the critical loading displayed well-dispersed hydrophobic domains that promote drug release. In contrast, formulations above this threshold formed continuous hydrophobic phases, which block dissolution. This insight explains why certain polymers successfully prevent precipitation while others do not.
What are the future directions and implications for drug development?
Christian: Looking ahead, we believe this physics-based modeling approach can transform drug formulation projects. It provides precise predictions and explains why molecules behave as they do, offering formulators unprecedented molecular insight.
While our study focused on PVPVA64, we’ve already proven the versatility of this method with another cellulosic-based polymer, HPMCAS. We look forward to publishing those results soon.
This predict-first approach gives scientists the freedom to design formulations with confidence. By immediately understanding where a formulation will achieve an optimal release profile – and where it will fail – researchers can focus their efforts on the most promising candidates. Replacing trial-and-error experiments with reliable predictive tools allows scientists to test more ideas, reduce costs, and ultimately produce fewer but more successful formulations.
If you’re interested in learning how amofor can accelerate your drug formulation projects, please don’t hesitate to contact Christian for a personalized consultation!