• High sensitivity MAM for early-stage activities facilitated by nano-flow LC-MS. • Reproducible characterisation of critical quality attributes with high sensitivity from limited samples. • Low sample consumption makes the method ideal for early-stage process development. • Non-targeted processing for new peak detection (NPD) consistently detected spiked peptides down to 0.6 fmol. During early-stage biotherapeutic development, analytical methods play an important role in candidate screening, process development, formulation screening and stability determination. However, developing sensitive and robust analytical methods is challenging in the early stages as there is often insufficient product knowledge and limited sample amount. Here we present a high-sensitivity multi-attribute method (MAM) centred on nano-flow LC-MS/MS for the characterisation of mAb-based molecules. Using NISTmAb as a reference, run-to-run variation was examined, in terms of peak area, peak width, peak asymmetry, retention time precision, and mass accuracy. The ability to detect and quantify commonly observed product quality attributes (PQAs) was also evaluated for a wide range of injected digest amounts from 1 ng to 100 ng. LOD and LOQ were between 0.6 to 7.0 ng of injected protein amount (R² > 0.99), for a selection of low abundant PQAs including deamidation, oxidation, succinimide, and lysine content, following regulatory considerations on the implementation of MAM in QC. The nano-flow MAM approach was then applied as an in-process test to study PQAs and impurities of a chimeric IgG1 mAb. Low amounts of samples were digested followed by high resolution nano-LC-MS, Orbitrap-based mass detection, operated within a Code of Federal Regulations (CFR) Part 11 compliant data system for data acquisition and data analysis. Finally, the new peak detection (NPD) aspect of the method was also evaluated, as an in-process test for detection of low-level impurities. The results highlight the potential of nano-flow MAM for early-stage process development and its suitability for QC applications.
Millán-Martín et al. (Sun,) studied this question.