Skip to main content
Premium Trial:

Request an Annual Quote

Q&A: LabCorp's Russell Grant on the Challenges Facing Clinical Mass Spec


By Adam Bonislawski

Name: Russell Grant
Position: Strategic Director and National Director of Mass Spectrometry, Laboratory Corporation of America
Background: PhD, University of Swansea; Senior Scientist, GlaxoSmithKline; Technical Director, Eli Lilly; Director of Mass Spectrometry, Esoterix Endocrinology

Russell Grant is strategic director and national director of mass spectrometry for the Laboratory Corporation of America as well as the as clinical chemistry chair for the American Society of Mass Spectrometry.

As such, he's well positioned to speak about the obstacles hindering clinical adoption of mass spec-based proteomics, many of which were discussed this month at the Association for Mass Spectrometry's fourth annual Applications to the Clinical Lab meeting (PM 12/20/2012), where Grant served as a member of the organization's scientific committee.

This week, ProteoMonitor spoke to Grant about the current state of the field and what advances are needed to make mass spec-based proteomics viable for large clinical reference labs.

Below is an edited version of the interview.

What are some of the challenges that need to be overcome to get mass spec-based proteomics into large clinical reference labs like LabCorp?

Certainly one of the bigger challenges in terms of translation is the actual complexity of trying to run proteomic-type, SISCAPA-type workflows at any volume. It's really been shown in the hands of elegant scientists but … we're not talking post-docs and PhDs running mass spectrometry systems in [a large reference lab]. So the workflows themselves that require elegance and actually running things like capillary LC and micro-LC and multiplexing these proteins together are still really very much in the purview of extremely skilled scientists.

I think the second issue … is a broader issue regarding mass spectrometry in general.

[T]he beauty of mass spectrometry is its ability to provide selectivity. It's an intrinsic measurement of a molecule, in this case a protein or a peptide. [Taking parathyroid as an example,] one of the things we see in PTH is there are a number of other C-terminal clip forms. So there are a number of isoforms, there are a number of clip forms, there are different posttranslationally modified forms of proteins. So mass spectrometry gives you a lot more data, but what does the data actually mean? We consider that the data we're generating is more informative, but it isn't, actually.

What we, I guess, appreciate is that antibodies capture a particular epitope and an auto-analyzing technology gives you an average protein concentration or amount, but we don't necessarily know which posttranslationally modified forms are being collected — whether a C-terminal or N-terminal clip are in place; whether there is an amino acid substitution. We just get that something grabbed hold of this antibody and grabbed hold of this other antibody, and now we've got a measurement … it doesn't give us anything other than the average capture or measurement of what is often a highly heterogeneous protein that we're trying to assay.

Well, let's invert that equation for mass spectrometry, and PTH is a good example of this. We can discriminate a number of different forms of that protein with mass spectrometry, but what does that tell us? Does it tell us that those additional forms are biologically active or not? We don't know. We don't know whether those other forms of PTH, for instance, are biologically relevant. So mass spectrometry gives us a lot more data, but I think fundamentally we have to go back to clinical trials … or [to] looking at those markers in well-constructed clinical situations, to determine what that extra data means so that the data can be converted into actual actionable information. Because right now we're just measuring more forms of a protein. But what does that mean?

So the specificity of mass spectrometry and its ability to resolve actually puts the burden back on the basic researchers, to my mind, to provide the appropriate understanding of what those particular markers mean within the context of clinical disease and clinical disease management.

So is all the effort that's being put into developing clinical mass spec platforms by companies like Agilent and Bruker and SISCAPA Assay Technologies (PM 1/20/2012), for instance, putting the cart before the horse? Or are such platforms going to be necessary to actually do the validation and clinical work that is needed to show that these different protein forms and new markers enabled by mass spectrometry are bringing clinically useful information?

Well, I do think that the complexity of the existing [mass spec] systems is going to require some significant investment and some significant evolution.

Are there any particular parts of current mass spec workflows that are particularly problematic from the point of view of a large reference lab like LabCorp?

Every aspect of it. Sample preparation, sample concentration, injection, separation, mass spectrometry, data reduction — it all needs an overhaul in terms of [making] what is a research tool into a robust, engineered tool.

We look at a lot of potential clinical protein biomarkers, and the discovery and the proof of principle that is executed in research centers and academia is of such a limited quality that we can't even consider replicating the workflow. We have to reinvent the workflow, mindful of the fact that the tools they were using and the controls that were used in these studies to generate this potential biomarker hit do not meet the standard that we would need to consider for a validated assay in the clinic.

There's a gulf in terms of quality of analytical rigor that occurs in a clinically validated assay on LC-MS/MS systems or mass spec systems in a clinical diagnostic lab relative to the feasibility, proof-of-principle, [and] generic workflows that are executed in biomarker discovery. I'm talking about thousands of samples, accuracy and precision statistics, and guidance that is not followed.

[Researchers say,] "Hey you know what? We did differential proteomics, and we found some things that are different." Wow. And you're going to hide behind the fact that you think your analytical platform is relatively precise? Sorry, but that gets you not very far. [Y]ou have maybe some interesting candidates, and if the disease is … completely untapped or we have real problems in disease management, then we have to pragmatically look at the data you've created; understand the weaknesses in the generation of that data; and make an educated guess as to whether we can overcome those [and] what the cost is going to be to overcome those.

Biomarker discovery, to me, in terms of mass spectrometry has really just so underperformed. And there's a misunderstanding [that the effort to clinically validate markers will] come from industry. I can assure you that it isn't. Because there are so many opportunities. And we're not bottomless pits of money.

So you think that industry, companies like LabCorp, probably won't be where this sort of translational work gets done?

I just think it's naïve to think that limited proof-of-principle data for new sets of biomarkers are compelling enough for industry to just dive into the fold and resolve all of the challenges and concerns. Because, quite frankly, the studies are incomplete; their control in measurement is certainly not to a standard that we feel comfortable with just translating. So we have to assess the rigor by which the data was collected. We have to assess the particular disease with which the biomarker is supposed to fit. And we have to ask ourselves a fundamental question: How much time and money is it going to take us to do this? Because we get hit with a lot of opportunities, and very, very, very few of them have gone beyond, "I have an idea. It's proof-of-principle. I'll sell it to a big company and they'll just pay for it." I'm sorry, but we're not bottomless pits of money. We're judicious scientists with the care of patients very much tattooed across our skin, and we're very, very cautious in what we do.

And so what you've seen, if you take that full circle, is that the majority of protein assays that people are talking about are already described clinical markers. Insulin growth factor 1, insulin growth factor 2, insulin, PTH. The poster child for mass spectrometry as a killer application or a killer assay is thyroglobulin. That's [University of Washington researcher] Andy Hoofnagle's work (PM 10/22/2010). It's working against autoantibodies, which is a known problem in immunoassay workflows for measuring thyroglobulin. That's a real problem, and [mass spectrometry is] a real solution.

The issue with the thyroglobulin [mass spec assay] right now is that, in my mind, we don't have that in a form where we could consider translating over from the existing workflows to mass spectrometry wholeheartedly because of the scale necessary to do that. It's so vast that there just aren't enough quality people and those workflows haven't been retooled or reengineered to deal with the volume of specimens that we would like to run.

Do any of the efforts that firms like Agilent have undertaken to automate and improve the throughput of mass spec-based protein assays hold potential in your opinion? What is needed to make an assay like the thyroglobulin assay practical for a company like LabCorp?

First, we need intact labeled proteins to control for the digestion of the protein. We need to get away from nanoLC and nanospray-type workflows into workflows where the separation modality is robust. We need more automated forms of doing the reduction and digestion. We can't use microliters of this and tiny volumes of that in a clinical setting. You have to automate it.

The robustness of the LC separation is part of the problem, the speed … of the separation. Andy [Hoofnagle's] separations are fabulous, but they're 15-, 16-, 17-minute retention times. Analytically that's a great research tool. Clinically, we couldn't buy enough mass spectrometers to deal with the sample volume if we're running capillary type flow rates. It's just not happening.

So, some of the moves Agilent is taking forward are good. I do like the design of their [6490] triple quad. I like their LC pumping systems. But they're a manufacturer. Are they going to take the lead on a [thyroglobulin] assay, for instance? I don't know.

What do you think of their work using the RapidFire system to try to eliminate the need for chromatography in some SISCAPA-type assays?

We'll believe it when we see it. And it's not proof of principle that we want to see. We want to see a fully validated assay, analytically and clinically. Because you can't just say this thing has promise. In the end, you have to have it validated to a standard.

Do you think efforts to develop such fully validated assays and platforms for running them are progressing well, or do you feel they have stalled?

[A] few laboratories and highly skilled people are taking appropriate steps forward and almost exclusively trying to improve the utility of existing clinical markers with mass spectrometry. You only have to look at the … relatively small list of protein assays that are commercially available with mass spectrometry measurement, and they are all proteins that we've known about for a long time.

Again, I want to reiterate that there is a disconnect between the discovery mode mindset and the people who really are extremely and exquisitely skilled in clinical diagnostic laboratories — particularly those who have brought highly complex mass spectrometry assays to the market and/or to the clinical utility endpoint.

There's a belief that, "if I build it, they will come." I'm sorry, but if your house has weak foundations, we ain't coming. It is a unique person who can take the research tools that we have and provide a high-quality, high-throughput, clinically actionable assay, particularly in the proteomics arena, because the workflows that … have been developed in discovery do not translate straight into the clinic. They're just not industrialized. They're not robust.

Have topics you'd like to see covered in ProteoMonitor? Contact the editor at abonislawski [at] genomeweb [.] com.

The Scan

Genetic Risk Factors for Hypertension Can Help Identify Those at Risk for Cardiovascular Disease

Genetically predicted high blood pressure risk is also associated with increased cardiovascular disease risk, a new JAMA Cardiology study says.

Circulating Tumor DNA Linked to Post-Treatment Relapse in Breast Cancer

Post-treatment detection of circulating tumor DNA may identify breast cancer patients who are more likely to relapse, a new JCO Precision Oncology study finds.

Genetics Influence Level of Depression Tied to Trauma Exposure, Study Finds

Researchers examine the interplay of trauma, genetics, and major depressive disorder in JAMA Psychiatry.

UCLA Team Reports Cost-Effective Liquid Biopsy Approach for Cancer Detection

The researchers report in Nature Communications that their liquid biopsy approach has high specificity in detecting all- and early-stage cancers.