Name: Laura Schnackenberg
Position: FDA Staff Fellow, 2003 – the present
Education: PhD, University of North Carolina at Chapel Hill, 2003
This month’s issue of the journal Pharmacogenomics features a special focus on the emerging field of metabolomics — the omics cousin to proteomics and genomics that concerns the pool of metabolites in a cell or tissue.
In an editorial accompanying five reviews of research on different topics in metabolomics, Laura Schnackenberg, a staff fellow at the Food and Drug Administration, writes about the simultaneous use of metabolomics, genomics, and proteomics to evaluate health and to investigate drug efficacy.
In fact the FDA has recently acknowledged the likely importance of these three distinct disciplines and has begun allowing drug makers to submit metabolomic and proteomic data as part of the agency’s voluntary genomic data submissions program (see related story, this issue).
Pharmacogenomics Reporter spoke to Schnackenberg this week to find out how metabolomics might contribute to pharmacogenomics. But one caveat: Schnackenberg differentiates between “metabolomics” and “metabonomics,” which she said refers to a special subset of metabolomics that deals with the response of living systems to drug toxicity or disease.
How do you see metabolic profiling fitting together with more familiar omics technologies in personalized medicine?
One of the advantages of metabonomics technologies is really that it has phenotype information in it as well. Previously with genomics and proteomics, just looking at the genotype [for example] may not necessarily be the best thing to predict whether somebody will respond to a therapy, or whether they’re going to have a toxic drug interaction.
What’s been found is that it’s not only the genotype, but the phenotype, and contributions to the phenotype come from health status and nutrition. We’re a very diverse population, which is why we need to take this all into account in moving toward personalized medicine.
So, the phenotype information that you can get from metabonomics is really going to be very important in determining what course of action is best for one person, versus someone else who has a different diet or maybe other pre-existing conditions that need to be taken into account as well.
One of the ways you can do this is — looking at each individual person — you can get their metabolic profile, monitor this over time — over the course of therapy, or just as a disease progresses — how the metabolite profile is changing, and if there is disease, are you bringing that profile back toward their control state prior to the intervention?
Are there particular diseases in which you can more accurately treat a person using combined information?
I think certainly you’re going to need to combine those technologies in any case, just to get a full understanding of what’s going on. Metabolomics is going to be the downstream markers of any changes in the genes or the proteins, but to really fully understand what kind of mechanism are in operation as a response to the disease, you really need to have a combination.
One of the examples in one of the review articles in this [current Pharmacogenomics] issue is that they’ve been looking at therapeutic targets for cancer. And so, by profiling not only just the metabolites, but genes and proteins, they can figure out what kind of targets they need to make for particular cancer drugs, and what’s going to be the best to target a particular gene or protein level.
What might metabolomics add to the equation, as far as what drug targets to attack?
It’s going to add a lot, because it’s going to tell you the phenotype information. One of the [other] advantages, too, is that getting information about metabolites is a lot easier — the technologies are a lot quicker — than for either genotypes or protein levels. So, you can kind of use the metabonomics to go back and start looking at those biochemical pathways.
For instance, you may see that a particular gene is upregulated or downregulated, but you don’t necessarily know what the end-path change is going to be. So, by looking at the metabolites, you can go back and actually trace it back to see what genes are being turned on or off at a particular time. Metabolomics will really help, I think, to develop better therapeutic targets.
Is there a way to incorporate that method into personalized medicine?
Oh, absolutely, and one of the goals of this technology is really to move into the clinic, because you can look at basic biofluids that are easily collected from a person, and look at those patterns of biomarkers that are changing.
Metabolic profiling has been used for a number of years to study various amino acid diseases, and certainly it will be widely applicable in the clinic, in terms of diagnosis, in terms of looking at how a patient has responded, and that kind of thing.
Is anyone actually using it in drug development for a personalized treatment?
There is at least one pharmaceutical company that is using it in some initial drug trials, and using that information to guide treatment and looking at how these specific biomarkers for whatever it is they’re looking at are changing in response to a drug.
Is there any reason to think that a metabolomic test might soon be paired with a particular drug treatment, analogous to the Herceptest/Herceptin model?
At this point I can’t say ‘yes’ or ‘no’ on that. Metabolomics is still in its infancy, and the [US Food and Drug Administration] is really just now trying to understand the data, and working with pharmaceutical companies to try to figure out how they will submit this kind of data to the FDA. So, right now the agency is just trying to process it and figure out all the data, so I can’t say one way or the other down the line whether it will require a certain test prior to [treatment with a specific drug].
Is the FDA taking voluntary data submissions for metabolomics?
Yes, they’re starting to. It’s just starting right now.
Rick Beger, who is my supervisor, has really been key in working on that. But they are getting ready to start requesting voluntary submission of data from metabolomics.
Through the so-called VXDS program?
It’s the same coordinators that were working on the VGDS—they’re working with Rick on getting this going as well. It will be a very similar program to the VGDS.
Have they taken any submissions?
I’m not aware of that. Through the VGDS, there was at least one that did include metabonomics data on it, but so far, I don’t think we’ve received any submissions just for metabonomics data. It’s just starting to be requested.
If you had to bet, when would you expect metabolomics to make its way into the clinic?
I think it’s definitely reasonable that it could be in the clinics within five to 10 years.
What’s another main use for metabolomics in drug development?
It should be possible to pick up markers of renal toxicity of liver toxicity—liver toxicity is one of the major reasons that drugs are recalled from the market.
What people have been seeing, and what we’ve kind of seen here is that you can get specific patterns of biomarkers that seem to be related to different types of toxicity, as well as cardiovascular disease. There have been examples of that in the literature as well, where it has been used to diagnose that kind of thing.