(M0930-06-40) Increased Throughput for Regulated Bioanalysis with Catch-and-Hold Mode of HPLC Separation with High Sensitivity Mass Spectrometers

Purpose: LC-MS/MS systems become a de facto standard equipment for regulated bioanalytical laboratories. One of the bottlenecks limiting the throughput of LC-MS/MS based methods is the run time per sample injection. After each injection, the HPLC column is usually regenerated by flushing out unwanted components from biological sample extract and re-equilibration prior to the next injection. In this study, we present a novel “Catch-and-Hold” mode (Figure 1) of HPLC separations coupled to higher sensitivity mass spectrometers. In this mode, HPLC was operated in isocratic or shallow gradients with a short runtime (less than 2 minutes) per injection and therefore doubled or tripled the sample throughput on the LC-MS/MS system.

Methods: The LC-MS/MS system used for this study composed of a Sciex Triple Quad 7500 mass spectrometer with an OptiFlow Pro Ion Source, a Shimadzu Nexara X2 UPLC system equipped with a SIL-30AC MP autosampler. Apixaban, ondansetron, and 1-methylnicotinamide were selected as test analytes. MRMs of each analyte were optimized based guided optimization procedure of Sciex OS (Version 3.0.0). Apixaban was resolved with an Agilent Eclipse Plus Phenyl-Hexyl (50x2.1mm, 1.9 µm) column with isocratic elution (40:60 MPA:MPB, MPA=2mM Ammonium Formate and 0.1% Formic acid in water; MPB=Methanol) at flow rate of 0.4mL/min; Ondansetron was resolved with a C18 column (Waters, Acquity UPLC BEH C18, 50x2.1mm, 1.7 µm) with isocratic elution (19:81 MPA:MPB, MPA=2mM Ammonium Formate and MBP=Acetonitrile) at flow rate of 0.5 mL/min; 1-methyl-N-nicotinamide was resolved on a HILIC column (Waters, XBridge Amide, 200x2.1mm, 3.5 µm) with isocratic elution (40:60 MPA:MPB; MPA=1mM Ammonium Fluoride and 0.1% Acetic acid in water; MPB=Acetonitrile) at flow rate of 0.8 mL/min. Shallow gradient such as 55%-65%B within a 1.5 min window without equilibration was also tested. Owing to higher sensitivity of the new type of triple quadruple instrument, extracted samples were further diluted with reconstitute solutions resulting 100-800 folds of total dilutions. All three analytes in study were extracted with protein precipitations from human plasma K2EDTA or human urine. Usually a smaller injection volume, 2 µL injection was used. For column cleaning and regeneration, after over 500-1000 injections, the HPLC columns were rinsed with isopropanol offline with dedicated raising HPLC pumps and re-equilibrated with starting HPLC gradient before next use. A separated identical HPLC column can be used for continued LC-MS/MS analysis if needed without disrupting the LC-MS/MS analysis due to column cleaning, regeneration and re-equilibration.

Results: Apixaban and ondansetron in human plasma samples have been analyzed with established and regular LC‑MS/MS methods with a run time each about 4-5 min/injection. To demonstrate the feasibility of the “catch-and-hold” mode of HPLC separation, we modified the existing HPLC gradients to either an isocratic elution or a shallower gradient elution with a runtime of 1.9 min (Figure 1). The extracted samples were further diluted (in a total 500-800 folds) due to increased sensitivity of Triple Quad 7500. The cleaner extracts allow us to repeatedly inject many samples without performance degradation of the HPLC columns. For both apixaban and ondansetron, a batch size between 500-1000 injections were achieved without column cleaning and regeneration. Both standards and quality control samples met typical acceptance criteria (Figure 3). 1-methylnicotinamide is an endogenous biomarker in both urine and plasma and was selected to test the “catch-and-hold” mode of LC-MS/MS method. With a HILIC column and isocratic elution, a short 1.9 min LC-MS/MS method was setup to analyze a complex set of biological extracts with standards and quality control samples. A total of ~500 injections were performed with repeated injection of a set of 96 samples containinging both urine and plasma extracts. Consistent responses of QC and standard were observed during the course of the injections. Detailed comparisons between catch-and-hold” method with traditional LC-MS will be presented with Ondansetron as an example. There are no significant impacts on peak width, peak intensity (signal to noise ratio) and resolution, despite potential matrix buildups with “catch-and-hold” mode. (Figure 2)

Conclusion: With “catch-and-hold” mode HPLC separation, unwanted retained matrix components are retained and accumulated on HPLC columns without incur a wash-out over hundreds of consecutive injections. This new operation mode eliminates the need for washing and equilibration steps, resulting in shortened run time. The increased sensitivity of current LC-MS/MS system, coupled with additional sample cleanup techniques, allow hundreds of injections (batch size) up to 1000 injections possible with very short run time. Therefore, this approach of LC-MS/MS method setup can double or triple the throughput in a regulated bioanalytical laboratory.