Developing an analytical method can be one of the most rewarding jobs an analytical scientist can do, but it can also be one of the most complex and frustrating. To help guide your practical experiments and thought processes we spoke to Kean Woodmansey to benefit from his experience.

Kean is the Senior Global Market Development Manager for pharma at SCIEX. Before joining SCIEX Kean worked as a Study Director in bioanalytical method development for a contract research organization for several years. Here, he describes 5 key steps and his thought process for developing a new bioanalysis method.

1 – In silico research

As with many things in life, good preparation is key. I always started by looking at the physicochemical properties of the compound or compounds I was looking to develop an assay for. I had a license for Marvin Beans which allowed me to draw the compounds, obtain the monoisotopic and average masses, and garner key information on properties that could be used during chromatography and sample preparation such as the isoelectric point (pI), LogP, and LogD. The key info here is the charge on the compound at a given pH. This is useful for chromatographic development as neutral compounds are retained later by reversed-phase chromatography. It’s also useful for extraction development as a compound that can be modified by pH to be charged or neutral is amendable to ion exchange or mixed mode solid phase extraction (SPE). Such a technique would be orthogonal to typical reversed-phase chromatographic conditions. If the compound is known I’d need information on solubility, stability, and potential metabolism.

Our sister company, Phenomenex, has a selection of solutions specifically designed for the pharmaceutical industry, and a series of sample preparation guides, such as method development for SPE with steroids.

2 – MS method development

I’m showing my age now but having been trained on API 365 and API 3000 instruments, I always preferred to tune MS instrument parameters manually by infusion. I’d start with a solution at 1 µg/mL infused at 1 µL/min teed into the mobile phase flowing at 500 µL/min. Source temperature, collision energy, and declustering potential were always the parameters that were most ‘tunable’. SCIEX auto-tune options have become very sophisticated, so I’d recommend using this method now. Keep an eye out for adducts and compounds with the potential to multiply charge. If using a triple quadrupole system, charge state distribution can often be kept narrower by using 0.1% DMSO in the mobile phase.

3 – LC method development

Once all compounds have been optimized on the MS system it’s time to move on to the chromatography development. Even with several different types of stationary phases available, I’d still estimate that 90% of bioanalytical assays use a C18 column, and you’ll have a good idea if this phase is going to be suitable or not from step 1 as you’ll have a LogD plot across the pH range. You have a good chance of developing a successful chromatographic method on a C18 column if the LogD is above 1 within the usable pH range (typically pH 2.5-10.5). In practice, there’ll usually be a usable ‘pH window’ that can be used. Hydrophobic retention is increased by running compounds in their neutral state; so bases in base and acids in acid but again, in practice, for many compounds the use of acidic modifiers will yield good results. It sounds obvious given for MS to work, ions need to be in the gas phase, but ensure all solvents, buffers, and organic modifiers are volatile. Typical mobile phases for running a gradient might be: Organic – 0.1-0.5% formic acid in acetonitrile, Aqueous – 0.1-0.5% formic acid in 10 mM ammonium formate. If tweaking the pH or changing to another buffer, check that the buffer will actually buffer at the new pH!

Make a ‘non-extracted’ sample from your tuning solution diluted in a mixture of the mobile phase at the starting conditions of the gradient. Grab a fresh C18 column (50×2.1 mm or 100×2.1 mm) and run a ‘scouting gradient’ at 500 µL/min from 5% organic to 100% organic over 4 or 5 min before returning to starting conditions. Calculate column volume and ensure sufficient time for re-equilibration. Stable pressure helps to show this equilibration time is correct and the retention will drift in if not. Hopefully, you’ll see the peaks of interest somewhere within the time of the gradient. Adjust the flow rate and gradient slope to suit but for 1 or 2 compounds and internal standards, I’d aim for no more than a 3-minute cycle time. Modern columns can tolerate a pretty high flow rate with minimal loss in efficiency (as explained by the Van Deemter equation) and you can use this to speed up the cycle time.

There’ll be times when peak shape is bad or there is minimal retention using C18. When that happens, I’d often go directly to the farthest extreme of selectivity and work back. Something like a phenyl or biphenyl column using methanol instead of acetonitrile can give good results.

Demonstrate intra-well precision before moving on to the extraction development.

If you would like help selecting the best column phase for your analysis, this HPLC/UHPLC Column Phase Selection Chart from Phenomenex is a great resource, along with SCIEX LC-MS/MS Fundamentals.

4 – LC-MS/MS Troubleshooting

Troubleshooting is a massive topic and doing it justice would require separate sections for each topic above, so I’ll keep this at a high level. As far as LC-MS/MS troubleshooting goes, the best advice I could give is to always start at the mass spec and work backward through the LC system to the sample. Simple and common issues like blockages or leaks post-injection can be quickly fixed without too much trouble. Regular preventative maintenance and service is a must along with good ‘LC hygiene’. Ensure you use ultra-pure solvents that are fresh and keep an eye on the number of injections on the column, checking for changes in backpressure and peak shape. For the mass spectrometer, look out for contamination signs in your mass spectra; thorough cleaning of the ion source might be necessary if you see unexpected peaks or reduced sensitivity. I’ll say this again as it’s important, but consistent maintenance routines can prevent many common problems, so keep a log of your system’s performance to spot trends or anomalies early on.

5 – Summary

Yeah, a quick Google search will provide any number of quotes about being prepared, but genuinely, the upfront work you do before even entering the lab is crucial. Don’t skip the in-silico stuff and literature search, because there’s no point reinventing the wheel. Also, things will not go to plan. Often. Work through issues in a stepwise manner changing one variable at a time. If you don’t you’ll never know what it was that fixed the issue. Hard to do when you’re reaching the end of the development period, study samples are coming in 2 weeks and there’s still a validation to do. I never said it wouldn’t be a bit stressful, and I’m sure many of you reading this can relate.

Learn more about SCIEX solutions for pharmaceutical assays here > Biopharma / pharma research

Join the SCIEX community and subscribe with us > Here