Q: Can you provide some details on the work you do and the techniques you use?

A: The Biomarkers Core Laboratory of the Irving Institute for Clinical and Translational Research at Columbia University is a large, highly specialized clinical chemistry lab that aims to serve the needs of the entire research community of Columbia University. A lot of the work involves using manual and automated chemistry- and immunochemistry- based assays that use different technology platforms. We also have a large mass spectrometry (MS) facility that focuses on small molecules. One part is focused on targeted metabolomics, and the other on drug assays such as absorption, distribution metabolism, and excretion (ADME) and Phase I studies. I also direct another lab, the Clinical Pharmacology and Toxicology Laboratory at Columbia University Medical Center, which is completely focused on applying LC-MS-based assays for therapeutic drug monitoring (TDM) in support of patient care. The Biomarkers Core lab uses both GC-MS (gas chromatography-MS) and LC-MS (liquid chromatography-MS), although recently we have been using more LCMS- based assays. In the TDM lab we only use LC-MS instruments.

Q: What are the advantages of using LC-MS?

A: LC-MS is fairly easy to use. If you have trained people who know how to use the technology to set up and validate assays, then LC-MS can be a very powerful and sensitive method to serve many research demands. LC-MS can be used to set up and validate assays relatively quickly, which can be very advantageous, especially for small molecule research. In the TDM lab where we run a large number of samples, factors such as speed, precision, and sensitivity are important. Hence, we routinely prefer to use LC coupled with MS. However, if we need to use incredibly large volumes of sample or need a very rapid turnaround time, then automated immunochemistry-based assays are preferred. These automated assays, however, do show cross-reactivity, often with metabolites of the drugs. Good LC-MS assays tend to be more specific and selective for the compound of interest, because they are not hindered by this cross-reactivity. However, there are problems with LC-MS, too, and you should have people who know what they are doing. I am very fortunate because in both of the laboratories that I direct, all of the people are outstanding.

Another reason for using LC-MS or HPLC-based assays is that you can develop the assays yourself fairly easily. Hence, you are not waiting for a manufacturer to develop an assay, which is very favorable in a research setting. This is true for preclinical, translational, and clinical assays, although the level of validation and quality assurance increases in that order, ultimately in our case adhering to GLP-guidelines for assays in support of Phase I studies. In the case of lab-developed LC-MS assays in support of patient care, you have to follow another set of regulatory and quality guidelines. In our case, being in New York State, we have to work according to the quality assurance system and continuously pass the tests outlined by the State Department of Health Quality Control program, the so-called Clinical Laboratory Evaluation Program (CLEP). This system guarantees a certain quality of our lab-developed assays.

Q: Are you seeing a shift away from GC-MS?

A: We have always used chromatography in our labs, starting with TLC (thin layer chromatography) and HPLC (high-performance liquid chromatography). GC-MS was the first MS-based assay to be used, and today we use mostly LC-MS. We are seeing a transition away from GC-MSbased assays, because LC-MS is easier to use, and we can measure more compounds without using any derivatization. However, GC-MS can be more useful, e.g., for very volatile compounds and for some organic acids. In our core lab we typically implement methods that have been published in the literature. If there is nothing published, then we develop our own methods.

Q: What is the main challenge when working with LC-MS?

A: Matrix effects can be a huge problem for LC-MS. For electrospray ionization MS, which is what we mostly use, the Achilles’ heel is electron suppression or enhancement, where the signal is decreased or increased by the endogenous compounds, and that can vary by sample or by patient. One way to tackle the problem is to use a stable deuterated isotope of the compound as an internal standard, but those aren’t always available, or they are sometimes incredibly expensive. Sometimes people use another compound that elutes at a different retention time than the internal standard, or they try to correct the problem without using any internal standard. Matrix effects are probably one of the most challenging issues to tackle when working with LC-MS.

Q: What innovations in chromatography have you seen or expect to see?

A: UHPLC (Ultrahigh-performance liquid chromatography) is definitely an improvement over traditional HPLC. It has improved chromatography and saves significant time, because the run times are much shorter. There have also been specific improvements to instruments from various vendors for specific compounds. For instance, the newer version of one of the instruments that we have from Waters Corporation has significantly improved sensitivity, specifically for steroid analyses, compared with its predecessor. We have a number of triple quadrupole MS instruments from various manufacturers. The new, improved models often have the sensitivity and ability to quantify compounds consistently without any drifts, and that is a big advantage. The downside is that these high-end instruments also tend to be very expensive.

Q: Will chromatography ever be used routinely in a clinical setting?

A: Chromatography is already in routine use for clinical applications. But it is important to note that chromatography itself becomes less important as the detection, which in our case is MS, gets better. With the new, fast instruments, you can very quickly run samples using a precolumn, and you can hardly call that chromatography. The art of chromatography is diminishing, even though you will always need it to some extent. It will probably never disappear completely, because we will always need some kind of sample cleanup.

Q: How important is it to impart the right training to people using chromatography?

A: For research applications, there are a lot of people who have experience with LCMS and can train others. The technique can certainly be learned in a couple of months. It is not rocket science. Anyone who is skilled can learn to use it, and the problems that are typically encountered are more on the chromatography end, not the MS end. However, in a method development lab, you do need more knowledge about LCMS to be able to better analyze the sample. So you need at least one highly experienced person who can do everything and train others as needed.

There is a real challenge if you do LC-MSbased assays for patient care, because here in New York State, for our TDM lab, you do need a licensed medical technologist to do the LC-MS analysis. Most programs for medical technologists offer limited exposure to and experience with LC-MS; hence, finding a person with such a background is difficult. We are trying to deal with this problem by collaborating with different local institutions and trying to partner with the medical technology schools in the tristate area to offer different programs. The New York State Health Department is accommodating us to some extent by giving out limited licenses to people who are very experienced with LC-MS, who are then allowed to run some patient samples. Getting the right people trained for this application is a big problem.

Q: Are there any other challenges that you would like to make lab managers aware of?

A: Setting up service contracts with vendors is another important issue. These GC-MS and LC-MS instruments all have very expensive service contracts, and depending on your need, you need to make sure to include these costs in your budgets. These contracts can be a very big part of the overall costs.

Q: What improvisations are you looking for?

A: Having a smaller instrument or one that can be more easily automated for clinical applications is an obvious need. Being able to link the instruments to the hospital’s existing laboratory information management system to be able to see and download results easily will be very helpful, especially for clinical use.