Name: Shelly Gunn
Title: Associate medical director of clinical genomics, PathCentral
Education: PhD, genetics, UTHSCSA Graduate School of Biomedical Sciences, 2002; MD, UTHSCSA Medical School, 2002
PathCentral, an Irvine, Calif.-based firm that offers cancer testing services, recently named Shelly Gunn as its associate medical director of clinical genomics.
In her new role, Gunn, who is also medical director at CombiMatrix Diagnostics, will help PathCentral implement "tech-only" array comparative genomic hybridization services for hematological malignancies, meaning that PathCentral will process the samples, but the data will be delivered to pathologists for their own interpretation.
Gunn has been applying CGH in the cancer setting for a decade. In addition to her role at CombiMatrix, she is also the founder and medical director for Oncopath Laboratory at the START Center for Cancer Care in San Antonio, Texas. Before that, she was an assistant clinical professor in the department of pathology at the University of Texas Health Science Center, San Antonio where she set up a pathology lab offering array CGH services (BAN 5/2/2006).
Her current translational research interest is developing array CGH-based clinical assays for scanning tumor genomes in hematological malignancies and solid tumors. BioArray News spoke with Gunn this week about her new appointment at PathCentral and the state of array-based cancer testing. Below is an edited transcript of that interview.
What does your new job at PathCentral entail?
PathCentral has been in the process of bringing CGH up for clinical use, but my job with them is to get it launched as a tech-only service. Their specialty at PathCentral is hematological malignancies. They have full flow cytometry capabilities and my goal for them is to have the best hematological array CGH available.
Building a tech-only program doesn't just mean designing the arrays and getting them optimized, it also means being able to report the data to a pathologist who may or may not have a lot of experience reading array CGH. It is very doable and as pathologists we learn new tests routinely, but you have to take the array and put it into the language [such] that a pathologist who probably didn't even go through a molecular path residency can interpret the data. We are going to be working very hard to make sure that array CGH is a routinely available test to our pathologists and that they feel comfortable interpreting it. By our pathologists, I don't just mean those who are working with PathCentral, but pathologists who are also ordering flow cytometry or immunohistochemistry or fluorescence in situ hybridization … should just be able to order CGH as a tech-only service.
How is that different from what CombiMatrix offers?
The strength of CombiMatrix is in the solid tumor arrays. We use those routinely for our patients here at the START Center through CombiMatrix's tech-only service. And then we have the Sequenom platform in our laboratory, so all of our patients get full chromosome analysis by CGH and full multiplex oncogene mutation analysis by Sequenom.
I like CombiMatrix's BAC array for solid tumors. It is a very robust platform that works well from the DNA that we get from FFPE tissue, which is about the only tissue you can get for the solid tumors that we see, the main ones being brain, breast, melanoma, lung, colon — they all come to use in tissue blocks. The DNA is degraded and not as easy to work with and so the BAC array that CombiMatrix offers is a fantastic way to look at copy number changes in solid tumors. PathCentral, though, will be offering the oligonucleotide array and that is fine for hematological malignancies. And their strength is going to be that they have flow cyometry right there on site. You really have to do flow before you can do any kind of CGH for hematological malignancies.
Why is PathCentral bringing CGH in house? Are its competitors offering CGH?
There are not as many as I would have thought, but FISH works well for most hematological malignancies and there aren't the targeted therapies yet for hematological malignancies that there are for solid tumors, so I think that the drive and the motivation to run more CGH assays seems to be in the solid tumor world. Oncopath Laboratory is running [them] routinely for solid tumors at the request of our clinicians. For hematology, we have to go out and say, 'Don't you want to look at the whole genome?' But for PathCentral, it is good for them to have the capabilities, but it is not going to be the core technology at that lab as it is at CombiMatrix.
Will you continue to work with PathCentral once the service is up and running?
Absolutely. They have a lot of projects ongoing. I am helping them with their genomic business as they build that side of their capabilities. They have some innovative ideas of how to use CGH ultimately for hematological malignancies. I am enjoying being in a planning and designing mode with them. Also, they are very good at information systems. The software side of CGH data is part of that design.
When you say CGH, what arrays are you using?
Both companies use the Agilent CCMC array and then I think that PathCentral will do some other designs, but I can't really speak about those yet.
Will you be offering arrays with SNP content too?
I think both [PathCentral and CombiMatrix] are going to be adding it. It is based on the clinical demand. For solid tumors, it is not important yet, but for hematological malignancies and developmental disorders certainly it is a capability that is going to be part of the arrays going forward. A big difference between PathCentral and Combimatrix is that PathCentral is exclusively interested in oncology and they are not focused on developmental disorders, whereas, pre- and postnatal array CGH is a big strength at CombiMatrix.
What is your opinion of the current generation of arrays?
I think they are excellent. The ones I use the most are Agilent and CombiMatrix's BAC array. There are those who say that BAC array is an outdated technology, it is not the array of the future, but it sure works well on solid tumors. And it is able to give us answers than an oligo array can't give on a degraded DNA sample. The oligo arrays I use now are loaded with cancer genes. We use the Agilent 180K because it provides adequate genomic information with enough content to cover the genes of interest but doesn't generate overly noisy data.
It seems that chromosomal arrays have gained more acceptance for use in constitutional cytogenetics than cancer cytogenetics.
I agree that CGH is certainly standard of care in developmental disorders whereas it is not in cancer. There are a lot of reasons for that. I would say the main one is that it is a diagnostic test for developmental disorders. For cancer, CGH is really hardly ever a diagnostic. It is a post-diagnostic, prognostic, predictive test. In hematology, it is not as essential right now as it is in solid tumors where we actually have the need for a post-diagnostic workup in a growing number of solid tumors, where we need to identify the biomarkers that clinicians can use to personalize cancer care according to the genomic profile of the tumor.
Have you been getting reimbursed for these tests?
Reimbursement is improving. We wouldn't be running CGH as routinely as we are if insurance companies weren't covering the service for some of our patients. The insurance companies are starting to take notice of the fact that tumor genome profiling can identify patients more likely to respond to chemotherapy before we choose a drug that may or may not have a therapeutic effect for a patient. The cost of the test is so much less than the cost of the drug that there actually is some growing support for some of this testing to be done before patients are given drugs that they may or may not respond to.
Have you thought of adopting next-generation sequencing?
All of us in the cancer world would like to be doing complete whole-genome sequencing of every single tumor we see. It is not routinely available yet. We have a study here at the START Center that has been funded called the San Antonio 1000 Cancer Genome Project where we are raising money from the community to bank DNA and sequence a thousand tumor genomes. We have already done some of them. We send them to Complete Genomics. The data that comes back is absolutely incredible. We would like to have that for every single cancer patient. But it is not covered by insurance yet, the bioinformatics are not there to make clinically relevant sense out of the thousands of biomarkers we get back, so it is not quite ready for prime time, but it is certainly is something we would like to do.
Copy number changes are very well addressed by whole-genome sequencing. Eventually, I think a lot of cancer patients are going to have a complete genomic profile of the tumor done by whole genome sequencing, but because we are not there yet we are using surrogates. The surrogates for us are CGH; we get good copy number info that way. Another surrogate is the Sequenom platform, which analyzes hundreds of mutations on a single chip for one patient. So we get for our patients now as complete a genomic picture of their tumors as we can by using the technologies that are available today. But there is certainly a demand for whole-genome sequencing. We have several motivated patients who have paid out of their pocket to have whole-genome sequencing done and we do our best to interpret the data when it comes back, but that is the future.