This interview with Jean-Claude Marshall, Director of the Clinical Pharmacogenomics Lab at Pfizer, first appeared in Issue 2 of Front Line Genomics Magazine.


As a prominent member of the genomics community at one of the world’s largest drug developers, Jean-Claude Marshall gives us his thoughts on where pharmacogenomics needs to focus, NGS and the FDA’s draft guidance for LDTS.

Developing personalised medicines is one of the major applications of genomic research. Pharmacogenomics is the study of the role that genes play in drug response. Traditionally, this has focused on linking single nucleotide polymorphisms with a drug response. Now, with the birth and proliferation of whole genome sequencing, the scope of what pharmacogenomics can, and will, encompass has grown significantly. Bringing this to fruition will require attention in certain areas, particularly in the recording of phenotypic data through large scale clinical trials. Communication within the research community and regulatory authorities is crucial to keep developing useful and relevant information. One thing is certain – pharmacogenomics laboratories are going to be even more invaluable in modern drug development.

FLG: What does the term ‘pharmacogenomics’ actually mean today? Now that you’re not just dealing with SNPs, it seems that the definition has expanded to encompass a lot more.

JCM: The term pharmacogenomics has indeed significantly expanded in meaning from the linkage of a single nucleotide polymorphism (SNP) with a drug response, to now encompass any genetic change related to both drug response and toxicity. This change has largely been driven by the increased use of next generation sequencing, particularly whole exome and whole genome sequencing of patients from clinical trials.

FLG: NGS is probably the technology that has really opened things up quite a lot. How big of an impact has it had on pharmacogenomics up to this point?

JCM: I would characterize NGS as just beginning to have significant impacts on the field of Pharmacogenomics. We have begun to see case reports of rare variants being described in the literature with clinical impact. Where NGS has already changed the field significantly is in our understanding of what constitutes a reference genome versus true variants. By that I mean we have much better understanding of variants at a population level than we did even five to ten years ago.

FLG: Are we still in a position where ‘more is better’ when it comes to producing data?

JCM: I think that the prevailing thought in the field is that more data is better, as long as you have a clear indication of what you are going to do with it. The generation of large amounts of sequencing data without any associated clinical data is much less useful. For example, there is currently a significant limitation on making phenotype associations with NGS data, compared to the progress that has been made in linking single SNPs or haplotypes with phenotype.

FLG: With more and more data being produced, the need for common standards grows. How tough is it to harmonize data across such a quickly growing field?

JCM: The rapid expansion of NGS has led to multiple different approaches for data analysis pipelines, some of which are slowly beginning to become standardized now. The major challenge that we still face is the integration of this genomic data into meaningful data sets, such as electronic health records. Even more important is to break down the data silo problem which has evolved across academic institutions, government and industry as a whole.

FLG: A lot of drugs fail in development due to safety issues. We’re starting to see some interesting work being published that could potentially help bring that failure rate down by expanding the predictive capabilities of data mining. Is bringing the cost of development down the big win for drug developers, or is there a bigger win out there than increasing efficiency?

JCM: Anything that increases our success rate in later stage clinical development will have a significant impact across all pharmaceutical companies. If there’s a way to predict for significant toxicity events, such as liver or kidney toxicities, before we reach first ‘in human’ studies, that would certainly be considered a significant advance. In addition, we need to remember that many drugs fail for lack of demonstrated efficacy or lack of improvement over current therapy. The promise of pharmacogenomics is to bring greater stratification of patients, creating enhanced efficacy or reduced toxicity, what we refer to as ‘Precision Medicine’.

FLG: We hear a lot of great things about the power of genomics, but progress can be frustratingly slow. Is there a risk that the clinical environment is too cautious to allow the field to really flourish?

JCM: Advances in scientific research are almost always a few steps ahead of clinical implementation. The cautious approach clinically is justifiably tied to a desire on the part of clinicians to base their decisions off of a solid base of scientific and clinical evidence. The competing drive between clinical caution and scientific drive for advancement has generally played out well. The area that may lead to significant delays in genomic personalized medicine may be the current uncertainty in the regulatory environment around these genetic tests. I would note however, that in the field of oncology, incorporation of pharmacogenomics has led to rapid demonstration of efficacy and drug approvals

FLG: Regulators are also in a difficult position. While their purpose is to ensure patient safety, they do seem to create a challenging environment for innovation at times. The FDA’s LDT plans are proving less than popular at the time being. Is there anything regulators could realistically do to help support innovation as well as protect consumer safety?

JCM: The question around the FDA’s draft guidance for LDTs is not an easy one. On one hand you have companies who are offering direct to consumer testing that does not necessarily require significant clinical validation. On the other, you have laboratories that are able to respond to scientific advances and offer testing rapidly due to the current CLIA oversight regime, such as the rapid deployment of Ebola testing. Striking a balance between the two may require repeated rounds of talks between the FDA and the clinical laboratories they are proposing to regulate. Neither side has historically interacted with the other and that may prove to be just as challenging as forging a balanced regulatory approach.

FLG: Who do you think is likely to feel LDT regulations most?

JCM: Clinical laboratories who fill niche markets, those who make early testing available in response to scientific advances, such as those offering NGS testing right now, and small clinical labs who don’t necessarily have the capability to interact with the FDA will feel these proposed regulations the most. That isn’t to say that the current draft guidance by the FDA won’t affect others. In fact, I think it’s clear that this draft guidance, if implemented in its current shape, may have broad reaching implications for the entire clinical testing field within the United States.

FLG: As well as being an expert in setting up regulated genomics laboratories, you also specialise in cancer. ‘Curing cancer’ is one of the real holy grails out there at the moment. With emerging genomic technology and its applications, are we in a position to realistically say that we have the tools to launch a devastating attack on cancer?

JCM: Every five or so years someone in the field comes forward and makes a bold declaration that they expect “cures” for cancer within the next few years. I think that we’ve finally begun to understand the fact that there aren’t just a few types of cancer, there are hundreds, perhaps thousands of types. There’s clear evidence that we are making progress on attacking cancer using these understanding, for example immune therapies and antibody drug conjugates. There’s still a lot of work to be done though, and I do have hope that we are making progress.

FLG: What’s the most exciting part of working at one of the world’s biggest pharmaceutical companies?

JCM: The breadth and width of the clinical trials that my laboratory supports is something that continually excites me. Being involved in trials from multiple disease areas, and being asked to support new clinical targets is something that is both energizing, and challenging. Being able to interact with people who are all experts in their realms and contributing to those teams is equally rewarding.

FLG: Prior to your work with Pfizer, you were at the CHI Center for Translation Research. Part of your work there was to engage with physicians to help drive genetic testing for solid tumour samples. What were the biggest obstacles to over come in getting front line practitioners bought in?

JCM: The education of clinicians around what molecular testing was, and how it could directly impact their patient treatments was always a challenge. There are multiple, sometimes contradicting, demands on a clinician’s time and resources which only adds to that challenge.

FLG: How do you think drug development and healthcare might change over the next few years?

JCM: I think that we are working during a time of great change in both the drug development and healthcare fields. The FDA regulatory oversight of LDTs will most probably continue to play out over the next few years, along with rapid advances in our understanding of genetic variants and the utility of whole exome and whole genome sequencing. There are many exciting opportunities available to us, although certainly those don’t come without risks. My hope is that we will begin to truly see the development of precision medicine, translating to direct improvements for patients across not only oncology but a range of disease areas such as cardiology and pain.

FLG: Is there any advice you would give to people just starting out in their scientific careers?

JCM: I think that we are working during a time of great change in both the drug development and healthcare fields. The FDA regulatory oversight of LDTs will most probably continue to play out over the next few years, along with rapid advances in our understanding of genetic variants and the utility of whole exome and whole genome sequencing. There are many exciting opportunities available to us, although certainly those don’t come without risks. My hope is that we will begin to truly see the development of precision medicine, translating to direct improvements for patients across not only oncology but a range of disease areas such as cardiology and pain.

FLG: Is there any advice you would give to people just starting out in their scientific careers?

JCM: This is something that is often said, but I think bears repeating: do what excites you and what you love. Find the area of industry, government, academia or a hybrid of all of those that challenges you and forces you to expand your thought processes, and go after that. Even failing at the right challenge, while painful initially, can be rewarding. 

More on these topics