An ultrafast and flexible liquid chromatography/tandem mass spectrometry system paves the way for machine learning driven in vivo sample processing in early drug discovery

• Published in: Rapid communications in mass spectrometry Vol. 35; no. 12; pp. e9096 - n/a
• Main Authors: Häbe, Tim T., Späth, Christian, Schrade, Steffen, Jörg, Wolfgang, Süssmuth, Roderich D., Bischoff, Daniel, Luippold, Andreas H.
• Format: Journal Article
• Language: English
• Published: England Wiley Subscription Services, Inc 30-06-2021
• Subjects: Automation; Chromatography; Data acquisition; Data points; Decision analysis; Flexibility; In vivo methods and tests; Liquid chromatography; Low speed; Machine learning; Mass spectrometry; Scientific imaging; Spectroscopy
• ISSN: 0951-4198; 1097-0231
• DOI: 10.1002/rcm.9096

Rationale
The low speed and low flexibility of most liquid chromatography/tandem mass spectrometry (LC/MS/MS) approaches in early drug discovery delay sample analysis from routine in vivo studies within the same day. A high‐throughput platform for the rapid quantification of drug compounds in various in vivo assays was developed and established in routine bioanalysis.
Methods
Automated selection of an efficient and adequate LC method was realized by autonomous sample qualification for ultrafast batch gradients (9 s/sample) or for fast linear gradients (45 s/sample) if samples required chromatography. The hardware and software components of our Rapid and Integrated Analysis System (RIAS) were streamlined for increased analytical throughput via state‐of‐the‐art automation while maintaining high analytical quality.
Results
Online decision‐making was based on a quick assay suitability test (AST), based on a small and dedicated sample set evaluated by two different strategies. 84% of the acquired data points were within ±30% accuracy and 93% of the deviations between the lower limit of quantitation (LLOQ) values were ≤2‐fold compared with standard LC/MS/MS systems. Speed, flexibility and overall automation significantly improved.
Conclusions
The developed platform provided an analysis time of only 10 min (batch‐mode) and 47 min (gradient‐mode) per standard pharmacokinetic (PK) study (62 injections). Automation, data evaluation and results handling were optimized to pave the way for machine learning based on decision‐making regarding the evaluation strategy of the AST.