Home » Q&A: Sharon Plon on Baylor College of Medicine’s First Year of Clinical Exome Sequencing
By Julia Karow
This is part one of a two-part interview. Part two will appear in next week’s issue.
Name: Sharon Plon
Position: Professor, Department of Pediatrics and Department of Molecular and Human Genetics; member, Human Genome Sequencing Center, Baylor College of Medicine
Chief of Cancer Genetics Clinic, Texas Children’s Hospital
Experience and Education:
Fellow in medical genetics, Fred Hutchinson Cancer Research Center, University of Washington, 1990-1993
Postdoctoral fellow, National Cancer Institute, 1988-1990
Residency training, internal medicine, University of Washington, 1987-1988
MD and PhD from Harvard University, 1987
SB degrees in chemistry and chemical engineering, Massachusetts Institute of Technology, 1980 and 1981
As a member of Baylor College of Medicine’s Whole Genome Laboratory management committee, Dr. Sharon Plon has been a leader in establishing Baylor’s clinical exome sequencing diagnostic test, which was launched last November.
She is also one of two principal investigators for Baylor’s grant under the National Human Genome Research Institute’s Clinical Sequencing Exploratory Research Project program (CSN 2/1/2012), which will explore exome sequencing for childhood cancer patients.
Since Baylor’s WGL started offering its clinical exome test a year ago (CSN 11/16/2011), it has received more than 600 samples and issued about 400 clinical reports.
At the Personal Genomes and Medical Genomics meeting at Cold Spring Harbor Laboratory last month, Plon gave a report of the lab’s experience after its first year of testing. Clinical Sequencing News caught up with her during the conference and the following is part one of an edited version of the interview, which provides an overview of the test. Part two, which will appear next week, will address how the lab handles interpretation and reporting for the test.
There has been a big upsweep in the last couple of months, so we have gotten several hundred samples just in the last two months or so. The awareness is clearly going up; I think that people are starting to hear about successes, certainly many of the physicians who have sent us samples have sent us more than one, and I think that as physicians are having diagnoses made, they are of course much more motivated to have other samples come in.
The other thing is, a lot of these patients are only seen periodically — often physicians will see them once a year — and as patients are coming back into clinic, they are now being told that this is a new test that’s now available, and they get ordered at that time.
Over 50 percent are from clinical geneticists ordering the test. The next largest group is pediatricians — some of these may have sub-specialty training — and then neurologists. Over 85 percent of the initial 300 samples were from pediatric patients, and about 15 percent were adult.
Samples came from all over the country. Clearly, Baylor-affiliated hospitals have submitted samples, but not the majority of the samples. Baylor is a large academic testing lab; there are many different hospitals that use Baylor as their genetic testing lab, and many of the same physicians are now sending samples for exome sequencing.
The most common diseases patients present with are neurologic phenotypes. By that I mean children who have severe learning disability, autism, epilepsy, or unexplained seizures.
I think the difference between our experience and others is that since we launched it as a clinical test, any physician who feels that it’s helpful to their patient can submit a sample. So, for example, the first of my patients I submitted was a child with severe leukemia who had unexplained toxicity to one of the chemotherapeutic agents, and we had a question whether that could be genetic in origin. It’s really up to the individual physicians, although clearly, parents of children with intellectual disability really want an answer, and a large proportion of those children have been otherwise undiagnosed.
We handle it as we do any other test. When the sample comes in, we go through an insurance verification. There are a small number that get held up at that point, but this test is really in line with the cost of other genetic tests — the institutional price is $7,000 — it’s non-invasive, and many insurance companies have covered it.
We focus on blood samples, although we will take DNA from fibroblasts or other tissues if that’s appropriate. We use a capture platform that is commercially available [from Roche NimbleGen] but that Baylor helped to optimize, so we have had a lot of experience with it in our research sequencing. We use Illumina as our sequencing platform; we started with a GAII for a very short period of time, then we switched to a HiSeq, then to a couple of HiSeqs, and now we are looking to use even newer equipment, the HiSeq 2500, for some of the indications where you really need a fast turnaround time. The sequencing is done to a quite deep coverage. We aim for 95 percent of the target at 20x or greater coverage, and our experience over the last 200 clinical exomes is that we normally hit around 97 percent.
In addition, we recently added the mitochondrial genome to our test.
We currently confirm positive results by Sanger sequencing. I think that over time, we will all become more comfortable with which mutations are correct 99.9 percent of the time, and which aren’t. And also, we do Sanger testing to confirm the parental inheritance, so we are going to do it anyway. For the majority of the mutations, we will probably continue to do it for the major findings in the report.
We also use a SNP array that was really introduced as a QC measure. It’s a way to confirm that the mutations that are read from the sequence match the genotype calls from the array. If we do see a large copy number change, we will report it out. But I think that patients should have a high-quality diagnostic array before having exome sequencing. We do not require it, but as a clinician, I highly recommend it — you get the result in a couple of weeks, it’s less expensive, and also, it can be very helpful. For example, if the patient has a deletion of part of a gene, and then the exome finds a new mutation in that same gene, you could make a diagnosis of a recessive condition with those two pieces of data together.
One of the things that varies among them is whether they do the exome on a trio or just the patient.
We have put a lot of work into using the academic resources at the medical school in the reporting. For example, I just got an e-mail yesterday from the lab, there was an unusual variant in a cancer gene, and they wanted my opinion about that variant, whether I think it should be reported. I think we are doing a lot of intellectual curation. We are really trying to use the resources of the school to provide the best information to the doctor, as opposed to some of the tests that provide longer variant lists upfront, for some of the physicians who want to do more of the interpretation on their end.
Our reporting is very much focused around what the patient’s problem was. So what we report out actually varies with the patient, whereas there are some systems where the report is essentially the same, independent of what the patient’s problem is.
Actually, very favorably. If you look at the early data using arrays as a diagnostic test, it was about 10 percent, and people were very impressed by that. Now we have better arrays, maybe it’s about 15 percent, in a child with, for example, intellectual disability. But Christine Eng, the lab director, actually looked at all of the tests we offer, and the majority are in the 5 percent to 20 percent range, most holding around 10 percent positive results. So we have been quite impressed.
And I think it’s important to realize that this is version 1. For example, we have now added the mitochondrial genome to our test. We and others are certainly aware that the exome doesn’t cover every single clinically relevant gene, so we are looking at using other chemistries to try to sequence those difficult-to-sequence exons and to generate a more polished exome. There is going to be constant improvement.
So the question about the remaining 75 percent of patients, is it simply if we did an even better job with the exome, we would find a mutation? A number of those patients may have methylation abnormalities, or they may have abnormalities in non-coding regions. And until we all just do more testing, we won’t have an answer.
That’s actually something we discuss quite frequently. I think that it surely could be ordered by a physician for a healthy patient. In our focused report, they would predominantly get incidental findings, autosomal recessive carrier status and some pharmacogenetic loci. There are other tests designed for healthy adults where the focus is more on pharmacogenomics and things like that.
I think we have had one or two healthy patients, and we have certainly talked about expanding that. And looking at, if we are going to focus on more of a screening-type test for healthy adults, what should we do to make a better test for that scenario?
It’s been an amazing year. At Baylor, there was a very large group of highly experienced people who came together and said, ‘Let’s get it done!’
I think the biggest challenge has probably been turnaround time, and trying to make sure that we adequately annotate and report out while dealing with impatient physicians who of course want a result for their patient. Because there are many MDs around the table, we all understand that feeling, but doing it right, especially when you are new, takes time.
Our turnaround time is about three months, and we have worked hard to get that down. Part of it is just having done more; we now have a much better database of frequent positives in the pipelines and things like that, we have gotten more feedback from the molecular diagnosticians of what annotation they need in the informatic pipeline to make their job easier in the end. That’s been, probably, the biggest back and forth, between the people doing the reporting and the people creating the pipeline to try to get as much information as possible into the pipeline so that their job is easier at the end.