Biomarkers can be defined and utilized for many purposes. There are disease biomarkers as well as mechanistic or pharmacological biomarkers. To reduce attrition and better understand a therapeutic, it is essential to be able to monitor the effectiveness of the therapeutic during drug development. Measurement of biomarker levels related to a particular disease state can help to understand the drug’s effect in-vivo. KCAS routinely develops and qualifies proprietary biomarker panels for our customers to study the pharmacodynamic response.
Most recently we have focused on developing and qualifying a number of biomarker panels. The choice of panels has been driven by collaboration with customers or based on research of more generic molecule classes and metabolism pathways. One of KCAS’ long-term goals is to have a range of biomarker assays that can be available as an off-the-shelf service for our pharmaceutical customers as well as more customized panels. The availability of multiple technologies has provided the ability to measure biomarkers ranging in size from traditional small molecule to protein panels such our MSD 54 plex assay looking at chemokines, cytokines, proinflammatory and angiogenesis markers.
Our focus within the LC-MS/MS team has been on endogenous small molecule biomarkers. Much of the complexity with these biomarkers comes from the large number that can be observed within a compound class. As an example, we have validated an assay for 20 amino acids (20-plex) that has been applied to support studies in metabolism and rare disease.
There are a number of considerations when developing an assay to monitor a class of biomarkers. Having multiple endogenous analytes means that the physiological levels can vary widely so calibration may have to be based on multiple curves or multiple ranges. The amino acids in the 20-plex assay are present in the body at different levels but fortunately they can be covered by variations within a single range all using the same ULOQ and an array of LLOQs based on detectability and endogenous levels. This impacts the number and level of QC samples required to cover that dynamic range. It is ideal to have 1 large set of STDs and QCs that covers the entire analytical range. Using a LIMS system such as Watson allows a one-time assay set-up for each analyte. This means that ranges can be set with unneeded concentrations deactivated for use across multiple studies.
The analytical procedure is relatively simple comprising clean-up of a 20-mL diluted plasma aliquot using protein precipitation followed by derivatization. The reconstituted sample is analyzed using UPLC combined with TurboIonSpray LC-MS/MS on a SCIEX API-6500. The “best” method would utilize individual stable-labeled internal standards for each analyte. In this 20-plex panel, fortunately, all the internal standard were available and could be incorporated into the assay. This can provide a more robust assay. This level of tracking may not be necessary or possible for all biomarker panels. In contrast, our glycosphingolipid panel uses a single internal standard per lipid class.
The derivatized extracts can have a deletorious effect on the LC-MS system and the performance for some analytes can degrade with decreasing cleanliness of the system. In addition, the assay is run at a very high backpressure. Both contribute to columns needing to be switched out on a regular basis. As experience in handling these assays has increased the team has learnt to recognize the warning signs for potential issues so they can be averted ahead of time meaning repeat analyses are minimized.
Other challenges of an assay of this complexity emerge during day to day operation, particularly during sample analysis when dealing with larger run sizes. Data processing is a challenge where integration is needed for 40 separate peaks and data processing for a single run can take ~2 hours. Regulated studies are run through our LIMS system but importing 20 separate files for processing & reporting each run also adds considerable time. There is a huge burden on downstream review so it’s important that any QC or QA reviewer is trained in the complexity of these runs ahead of time. We have recently implemented macros to check the cleanliness of blanks and carryover to help save some time.
Biomarker panels such as the amino acid 20-plex are designed in the knowledge that some analyses will not require all analytes. Having 20 analytes fully validated means that we can have a flexible approach to meet customers’ needs. Off the shelf assays can be available for any subset ranging from 1 to 19 amino acids depending on the necessary readout from planned studies. Some of the more abundant amino acids can also be run on a less sensitive instrument such as an API-5000. However, we do routinely run the full set of 20 analytes for some customers. Extension of panels to include other analytes of interest can also be achieved.
KCAS is interested in increasing our suite of biomarker assays. We are happy to develop and qualify custom methods or collaborate on more general assays with shared ownership.