Whether identifying metabolic dysfunction in a newborn baby or the levels of a drug in a person enrolled in a clinical trial, mass spectrometry can provide a window into how individual molecules pulsing through the body affect health.

While mass spectrometry is a very sensitive analytical technique, it is a complex procedure to perform. The traditional mass spectrometry workflow often requires a series of manual and complicated steps to prepare samples before they even reach the mass spectrometer. This means that only technicians with this specific expertise can perform it.

Through discussions with mass spectrometry experts and key opinion leaders, researchers in the clinical diagnostics arm at Roche wondered if they could find a way to make mass spectrometry easier, thereby increasing its access to more diagnostics laboratories and more patients.

“We’ve got a promising diagnostic technology. But there is a lack of automation, and it's too complex,” said Benjamin Lilienfeld, Lifecycle Leader Serum Work Area Systems at Roche. But, he added, “That's where we actually have expertise. We were automating other technologies in the past like PCR or immunochemistry.”

In what took years of research, the team at Roche developed a new bead technology called paramagnetic particles to automate the most complicated step in the mass spectrometry process: sample purification and preparation. With this new automation capability, clinicians and researchers can bring the power of mass spectrometry as a diagnostic to more patients, ultimately leading to better care.

How do your paramagnetic particles work to help automate the mass spectrometry workflow?

The complex part of mass spectrometry is actually not just the mass spectrometry measurement or the liquid chromatography that comes just before it, but it's the sample preparation step. We need to precipitate the sample, also called an analyte, and then centrifuge it to enrich it. We remove the remaining liquid and so on. Every analyte has different procedures to do that, and these are quite manual. Because of this, we couldn’t just copy and paste one procedure and try to automate it. We had to come up with another way to purify analytes. That's where these paramagnetic particles came into play. We have experience with these kinds of beads from our immunochemistry systems such as our Elecsys technology, which relies on immunobeads. If the technology worked there, we wondered if it would work in the mass spectrometry workflow.

We used two concepts. We developed paramagnetic particles that are specific for a certain analyte. We would capture the analyte on the particle, and then, because the particles were paramagnetic, we could fix them. Then, we washed the beads to get rid of everything that we didn’t want, and we would have the analyte in a rather pure form. After releasing the analyte from the particle, the analyte could go through liquid chromatography-mass spectrometry (LCMS). It also works the other way, depending on the analytes we were trying to purify. Instead, we could capture all of the stuff we didn’t want to have in the sample. We could take that away and keep the analyte. In the end, there was a rather pure analyte that ran over liquid chromatography for the separation part, and then we ionized them and brought them back into the mass spectrometer.

How have the paramagnetic particles affected the mass spectrometry workflow?

We can provide about 100 test results per hour. That's roughly four to five times faster than most of the traditional methodologies. But — I would like to stress — it's not just about the time. In the traditional mass spectrometry workflow, people run samples in batches because the different analytes have different extraction and purification protocols. The paramagnetic particles and the internal standards that we use allow researchers to load their samples as they come in. They don't have to batch anymore. This means that people can immediately run their samples when they come in, which makes the process even faster, and it also allows for standardization across different laboratories.

What were some challenges you faced in developing these paramagnetic particles?

As in any research laboratory, we started with small volumes. As we developed the purification process further, we then had to start thinking about how to scale that up to a production level — from a few milliliters to several liters. Our industry scale is around 100 liters. Not surprisingly, we ran into different challenges during this upscaling process. For example, the particles started to stick to one another. With certain weights and volumes, it gets more and more complex. We had to find ways of dissolving the particles nicely in a solution because we had to distribute them into reagent packs that have the different components needed to run the mass spectrometry tests. It's not as simple as one might think in the beginning. Luckily, with our experience from the immunobeads, we worked with our internal experts as well as external experts to come to the stage that we are in now where we have this whole process under control.

We’ve got a promising diagnostic technology. But there is a lack of automation, and it's too complex. 
- Benjamin Lilienfeld, Roche

What kinds of samples can researchers analyze in this mass spectrometry system?

We have a number of analytes that we're planning to launch over the next several years on our cobas Mass Spec system. In December, we got a CE mark (Conformité Européene) for the system and the first set of analytes, which include a number of different steroid hormones like testosterone, estradiol, and others. We will launch more steroids during the course of this year. In February, we launched vitamin D, and later this year, we plan to do the same for a number of therapeutic drugs including anti-epileptic drugs, antibiotics, and immunosuppressive drugs for transplant patients. Over the next few years, we also want to invest in more therapeutic drugs and a drugs of abuse panel. At the moment, our plan is to launch around 66 different analytes in the categories I just described over the next several years, but we also have additional ideas for more analytes in the future.

How do you envision researchers using this new automated mass spectrometry technology?

As an example, a lot of breast cancer patients receive endocrine therapies to suppress estradiol levels because it has been shown that if that hormone is suppressed, the chance of cancer recurrence is lower. People take these therapies over years and years, but they're not always 100 percent effective for every person. Therefore, doctors regularly monitor the estradiol levels in these patients’ blood. If the levels go up, the physician has other therapies at hand, but it's important to detect that rise early. To do that, doctors need a sensitive methodology, and mass spectrometry turns out to be very sensitive for measuring estradiol. Without the wide adoption of mass spectrometry capabilities in routine laboratories, mass spectrometry results might not be possible. Physicians might use other technologies that are good but might not be sensitive enough for the specific case of an increase in estradiol. There are other examples around monitoring immunosuppressive drugs after transplantation, antibiotic stewardship, and so on and so forth. There's a lot of use cases where this really makes sense.

What do you find most exciting about developing new mass spectrometry technologies?

By making this process easier to use, we can integrate it into the overall routine lab workflow. We can make it accessible to more laboratories, especially laboratories that do not have a mass spectrometry specialist. That means that mass spectrometry-based diagnostics are accessible to more physicians and therefore patients — and that is what this is all about. This is what really makes a difference for patients, and that's what really drives us.

This interview has been condensed and edited for clarity.