High-throughput chromatography has become a familiar tool in early bioprocess development, where parallel experiments can save time and precious material. But using miniaturized systems to generate data that influence late-stage process characterization, validation, and commercial manufacturing controls has remained a tougher proposition. According to a new study by Kamiyar Rezvani, a scientist at AstraZeneca, and his colleagues, that gap might be narrowing.
Researchers evaluated automated 600-µL RoboColumns operated on a Tecan liquid-handling platform against conventional, roughly 20-mL ÄKTA bench-scale chromatography models and manufacturing-scale operations for purification of a bispecific antibody (bsAb). The goal was to determine whether the two scale-down approaches provided equivalent process understanding or introduced scale-dependent biases.
The work covered three purification operations: lambda light-chain affinity, anion exchange, and cation exchange chromatography. The researchers first qualified both scale-down models against manufacturing data and then replicated multivariate design-of-experiments studies across the Tecan and ÄKTA systems.
The attraction of miniaturization is significant. High-throughput purification tools offer “parallelization and material savings” compared with traditional bench-scale experiments and can “enable practical end-to-end integration to automate entire experiment workflows,” Rezvani and his colleagues note.
The results strengthen the argument for extending those benefits deeper into bioprocess development. RoboColumn scale-down models showed “excellent agreement with manufacturing scale at target operating conditions” and aligned with bench-scale results from multivariate studies, according to the authors. Perhaps more important for process characterization, the practical effect of differences between the two models on the resulting control strategy was “negligible,” wrote Rezvani and his colleagues.
That does not mean that microscale chromatography behaves identically to larger systems. Smaller working volumes, offline fraction analysis, intermittent liquid delivery, and differences in flow behavior can introduce variability. In the study, step yields and product column volumes “consistently varied between” scale-down models (SDMs), the researchers point out, although product-volume offsets were highly reproducible across broad process parameter ranges. As the authors emphasize, a “well-characterized SDM does not need to produce data that is identical to manufacturing scale” to be suitable. Instead, differences need to be understood, their risks assessed, and appropriate offsets or controls incorporated into the manufacturing strategy.
There are still practical limitations. The researchers recommend maintaining product-pooling consistency, minimizing evaporation effects, and ensuring sufficient product volumes for reliable analytical testing. For processes showing unusual or highly sensitive host-cell protein washes or protein recovery, they suggest additional microscale characterization or hybrid studies combining microscale and traditional bench-scale experiments. Plus, qualification remains essential. “Comparability to manufacturing scale must be verified through proper SDM qualification and scientific rationale,” the authors caution.
Taken together, the study shifts the question from whether microscale chromatography perfectly reproduces larger columns to whether its differences can be predicted and managed. The researchers reported that RoboColumn differences were “predictable both at target conditions and across varied process parameter ranges” for three chromatography methods commonly used in antibody manufacturing. For bioprocess teams facing increasingly complex biologics and pressure to accelerate development, that finding could move microscale chromatography beyond a screening tool—and closer to a platform for generating the process understanding behind commercial manufacturing decisions.
