Imagine that following a service call on your laboratory’s mass spectrometer, the retention time for your assay has shifted and the expected chromatographic peaks are missing. Do you as a laboratorian know where to start troubleshooting? While there is no substitute for experience, attending Sunday’s AACC University Short Course gave the attendees experience by proxy as the presenters, Judith Stone, PhD from the University of California San Diego Toxicology Laboratory, Alan Rockwood, PhD from the University of Utah and William Clark, PhD from the Johns Hopkins University School of Medicine shared their experiences with problematic assays.
Following a brief introduction on mass spectrometer components and general troubleshooting procedures by Stone, the three short course faculty members split the attendees up into three groups. In each group one of three presenters acted as discussion facilitators and used a case-based format, beginning each case with a statement of error, and showing the associated chromatograms and calibration curves as symptoms to involve the attendees in active learning.
All the cases shared were highly informative. One case detailed the method development of a benzodiazepine assay. The case description began with a nonlinear calibration curve for a lorazepam assay. The small group I sat in with was asked by Clark for ideas on possible causes for the nonlinearity. Some of the suggestions from the participants were saturation of the signal, or that the internal standard could be contaminating the lorazepam peak. Following discussion of these possibilities, it was revealed that the cause of the error was not due to instrument failure but rather due to the overlapping masses of deuterated internal standard and standards used for the calibration curve. This was a surprising outcome given that the internal standard should be 4 mass units greater than the unlabeled material. However, lorazepam contains two chlorides and about one quarter of chlorides atoms are the 37Cl isotope instead of the more common 35Cl. Thus the fact that some of the unlabeled lorazepam has the same m/z ratio as the internal standard was the cause of the nonlinear response.
Another case provided for discussion focused on an established voriconazole assay that presented with an unexpected shift in chromatographic retention time and signal intensity. Among many suggested causes, one of the group members suggested that he would check the liquid chromatography mobile phase, as it could explain both retention time and signal differences. Following further discussion and description of the troubleshooting steps taken, the solution to the problem was revealed to be an accidental switch of mobile phase composition.
Rockwood concluded the short course with some sage advice concerning troubleshooting. He suggested that mass spectrometry users should maintain a troubleshooting log detailing the problem and the steps taken to resolve it so that should the same event occur in the future, the solution could be quickly found.
As these cases illustrate, mass spectrometers come with unique challenges that require near Holmesian abilities to deduce the source of the problem. This interactive short course gave the attendees many key ideas that could be applied to troubleshooting their own troublesome assays.