In conversation with Hanns Christian Mahler, CEO and Co Founder of ten23.health, it quickly becomes clear where PAT reaches its practical limits. The challenge is not the generation of analytical data. The difficulty lies in consistently using that data for process control, embedding it within a regulatory sound framework, and translating it into stable, routine operations.

Mahler has seen both sides. After positions in large organizations, he deliberately moved into the CDMO world and founded ten23.health in 2021, with a focus on sterile injectables. The portfolio addresses demanding applications, ranging from formulation and process development to method development and the sterile manufacture of clinical and commercial products. Daily operations are shaped by stringent requirements for process control, combination products, container closure integrity, and reproducible filling processes.

From this practical background, PAT appears less as a project label and more as a logical consequence: develop stable processes as early as possible, manufacture under tight control, and limit post production testing to what is truly necessary. At the core is the question of how to establish robust correlations between process parameters and critical quality attributes through multivariate data analysis, and how to use these correlations sustainably in routine operations.

From Pilot to Daily Practice

“The decisive step begins with a clear use case,” Mahler explains. This does not refer to a laboratory demonstration, but to an application that is genuinely used for process control. A process owner in manufacturing is essential, not only in development. Without internal ownership and competence, any solution remains foreign to the organization. Early involvement of Quality Assurance and, where appropriate, dialogue with regulatory authorities are equally important. Only then does a realistic path emerge for implementing and validating models and systems.

This perspective helps explain why PAT has often been piloted in large organizations but not rolled out broadly. In practice, organizational factors tend to outweigh technical ones. When quality functions are involved too late, time is lost and trust must be rebuilt. The difference is noticeable when PAT is understood as a tool for process control rather than as an additional analytical technique.

Process Control Instead of End Testing

A concrete example from sterile manufacturing is one hundred percent in process control of critical attributes such as fill volume, including the possibility of adjustment. Concepts like this shift the focus. Instead of relying primarily on statistical end product testing, process capability is demonstrated and managed across all units. This directly affects traditional end product testing strategies. Where stable and comprehensive in process control exists, sampling plans can be justified differently than in an environment lacking comparable process data.

Mahler also observes how the term PAT has evolved in recent years. As soon as it becomes a catch all expression, skepticism increases. Ultimately, what matters is whether a parametric approach reduces unproductive end testing while simultaneously increasing process robustness. Anything else remains, at best, academic.

Traditional Medicinal Products versus ATMP

The contrast becomes particularly clear when comparing conventional medicinal products with Advanced Therapy Medicinal Products, or ATMPs. For traditional products, processes are often stable. Many batches exist, extensive data sets are available, and manufacturing generally functions even without PAT. In such settings, PAT is frequently used as an optimization tool, often in the context of real time release strategies.

In ATMPs, variability is inherent. In extreme cases, batch size may equal one. Continuous data acquisition becomes a prerequisite for maintaining control. At the same time, regulatory uncertainty is greater. Used correctly, this can facilitate innovation, because control and deep process understanding are not optional but mandatory.

Less Laboratory at the Line, More Control of the Line

One aspect that is often underestimated in PAT discussions is the objective itself. The goal is not to relocate the entire laboratory to the production line. The aim is to control the line so effectively that laboratory testing is performed where it truly adds value. This requires prioritization and a shared understanding of the purpose of each control. In practice, this is often the most difficult part, since many organizations traditionally operate according to a principle of maximum safeguard.

Looking Ahead and Suitable Formats for Exchange

For the coming years, Mahler expects PAT to continue its shift from a nice to have toward a foundation of robust manufacturing processes. Supply security, robustness, and reproducibility will carry increasing weight. PAT can serve as a decisive lever, provided implementation succeeds across the interfaces of manufacturing, Quality Assurance, and regulatory affairs.

An important question concerns suitable formats for professional exchange. General forums retain their value, yet the need for highly specific application examples is growing, tailored to disciplines such as sterile manufacturing, biologics, or cell therapies. The more focused the exchange, the clearer the requirements, limitations, and pragmatic solutions become.

In the context of emerging regulatory approaches, Mahler refers to the concept of regulatory sandboxes. These are controlled environments, supervised by authorities, designed for innovative technologies, decentralized manufacturing, AI supported release strategies, or continuous processes. Such approaches could help clarify framework conditions more rapidly. This is particularly relevant for PAT, as validation logic and regulatory acceptance would not only be discussed in theory but tested in practice.

Conclusion

The discussion makes one point clear. Today, PAT rarely fails because of missing technology. It fails when it is not implemented as an integrated system. A precise definition of objectives must come first, followed by clear organizational anchoring in manufacturing. Without early involvement of Quality Assurance and regulatory functions, the path becomes unnecessarily difficult. When these prerequisites are fulfilled, measurement values become true control variables, and PAT becomes tangible in daily production practice.

Definitions:

PAT – Process Analytical Technology
An approach for the continuous or timely acquisition and evaluation of process data, with the objective of improving process understanding, actively controlling manufacturing processes, and ensuring product quality. PAT encompasses sensors, data analytics, models, and organizational integration into development and manufacturing.

ATMP – Advanced Therapy Medicinal Products
Medicinal products for advanced therapies such as gene, cell, and tissue therapies. Characterized by high variability, small batch sizes up to n equals one, and specific regulatory and process related requirements.

CDMO – Contract Development and Manufacturing Organization
A service provider offering development, scale up, and manufacturing of pharmaceutical and biopharmaceutical products for various clients under regulatory frameworks.

QA – Quality Assurance
The function responsible for ensuring compliance with regulatory requirements, evaluating processes and data, and approving products and manufacturing concepts.

CCI – Container Closure Integrity
The integrity of primary packaging systems. A critical quality attribute for sterile medicinal products to ensure sterility throughout shelf life.