Expert Commentary on Formulation, Scale-Up, and Patient-Centered Design Found at CPHI Milan

As part of our coverage of CPHI Milan 2026, PharmTech spoke with experts about the challenges of turning complex molecules into medicines patients can reliably use. Their advice converged on a common theme: decisions about bioavailability, manufacturability, and real-world use should be made early and together, not one function at a time.

AI and Bioavailability

Andrew Parker, PhD, executive drug development consultant at Quotient Sciences, said that developers should assess a molecule’s chemical, physical, and biological properties together to shape a formulation strategy that fits the required clinical dose.¹ Mapping a compound on the solubility-permeability plot helps identify the enabling technology most likely to work, such as amorphous solid dispersions, lipid-based systems, or particle size reduction. Parker stressed that manufacturability and scale-up risk belong in the earliest stages; the same drug-polymer combination can behave differently when made by spray drying vs. hot melt extrusion. “Failing to plan for this early can force costly re-engineering later,” he told PharmTech. He added that artificial intelligence (AI) tools such as Bayesian optimization are already shortening formulation timelines compared with traditional design-of-experiments approaches.

Parker described AI adoption in formulation science as still early, with the technology augmenting scientists rather than replacing them.² When solubility and human bioavailability are poorly understood, he said, Quotient's Translational Pharmaceutics platform allows teams to defer dose-strength decisions until after first-in-human data are available. He also urged chemistry, manufacturing, and controls (CMC) and clinical teams to define what each stage actually needs instead of defaulting to commercial-ready formulations too soon.

Scaling Spray Drying

Rising molecular complexity and poor solubility are driving demand for commercial spray drying, according to Jason Bertola, executive director of Commercial Development for the Advanced Synthesis Platform at Lonza in North America.³ The biggest misconception, he said, is that the process scales linearly. Formulation-focused companies often underestimate drying kinetics, atomization, solvent management, particle collection, and secondary drying. Building in-house capacity makes sense only with a large, predictable pipeline; otherwise, a contract partner can offer faster access to expertise and commercial-scale assets. What matters, Bertola said, is “access to the know-how needed to scale successfully.”³

Designing for Real-World Use

Ron Scarboro, CEO of Azurity Pharmaceuticals, argued that medication non-adherence is a design problem.⁴ Teams should ask early who will take a medicine, where, and what might make that difficult, and treat barriers that surface after launch as design input rather than a reason to add patient reminders.

Reformulation must balance usability with performance, Scarboro said.⁵ A version that removes a fasting requirement, for example, must still meet the same standards for exposure, safety, and efficacy. “Convenience matters, but only if the medication continues to do its job,” he said. He recommended that one party own a product’s entire pathway, from unmet need through development, approval, supply, and patient use.

Scarboro said packaging and delivery choices should reflect where a product is stored, who administers it, and what each market's infrastructure can support.⁶ Rather than relying on patents and approvals as measures of success, he urged teams to track complaints, administration issues, product availability, and, where reliable data exist, adherence and persistence.

The hardest development problems, from poor solubility to poor adherence, may be best solved by connecting formulation, manufacturing, and patient insight from the very first decisions.

References

1. Parker A, Haigney S. CPHI Milan 2026: AI and bioavailability, part 1. PharmTech. Sept. 14, 2026. https://www.pharmtech.com/view/cphi-milan-2026-ai-and-bioavailability-part-1
2. Parker A, Haigney S. CPHI Milan 2026: AI and bioavailability, part 2. PharmTech. Sept. 17, 2026. https://www.pharmtech.com/view/cphi-milan-2026-ai-and-bioavailability-part-2
3. Bertola J, Haigney S. CPHI Milan 2026: the demand for commercial spray drying. PharmTech. Oct. 8, 2026. https://www.pharmtech.com/view/cphi-milan-2026-the-demand-for-commercial-spray-drying
4. Scarboro R, Cole C, Zubulake Z. CPHI Milan: medication non-adherence is a design problem. PharmTech. Oct. 1, 2026. https://www.pharmtech.com/view/medication-non-adherence-is-a-design-problem
5. Scarboro R, Cole C, Zubulake Z. CPHI Milan: why end-to-end pathway ownership matters. PharmTech. Oct. 5, 2026. https://www.pharmtech.com/view/why-end-to-end-pathway-ownership-matters
6. Scarboro R, Cole C, Zubulake Z. CPHI Milan: packaging and delivery decisions must reflect real-world care. PharmTech. Oct. 7, 2026. https://www.pharmtech.com/view/packaging-and-delivery-decisions-must-reflect-real-world-care