Learning and Growing with In-House Testing

An interview with Wei Song

How – and why – does your institution conduct precision oncology testing?
Precision medicine is a newly evolving discipline and, to fully realize its potential, all adequately sized institutes should be able to provide in-house genomic profiling for tumors from both tissue and plasma samples. I always think as though I’m running a startup company, so my number one consideration is the customer – who are they and what do they need? Our customers are oncologists and they need genomic profiling tests to enable their patients to benefit from novel precision oncology treatments.

Also, I believe molecular diagnostics is the future of pathology. So while centralization plays an important role currently, if we don’t practice in-house testing, we won’t be able to develop alongside the science and provide the best possible standard of care. That’s why we do as much as possible in-house – to support our clinicians and patients and to keep learning and developing.

Have you participated in a move toward test centralization?
The opposite, in fact. In our institution, all testing was initially sent to a central laboratory – but, ever since we established molecular profiling in our lab, it has come to us instead. Bringing our testing in-house was a real game-changer for the oncologists, for the institution, and for the entire community.

What do you think about test centralization and its promotion by large commercial laboratories?
First, it’s important to look at the central laboratories’ testing success rates. Some of them are unable to test as much as 30 percent of patient samples due to limited size and tumor content. Because we can test even very small samples in-house, our success rate is 98 percent. The second thing to consider is turnaround time. We provide results in three to four days. Most of the central labs can’t compete with that; it takes them one to two weeks – a huge difference. Third, pathologists and oncologists must work together to examine each patient case individually. Sometimes, we need to drill down to the detailed results and assess everything in clinical context. It can be extremely difficult to get detailed data from a commercial lab – and it might take up to a month. In-house, we’re ready 24/7. Any time a clinician calls us, we can jump on the computer and review the data with them virtually. This is hugely important for them – and it’s no less important for our development and for the research that drives modern medicine.

Precision medicine is in its infancy, so continual development is key and every case deserves a thorough investigation. If you look at the molecular testing report without also having the opportunity to examine the raw data and interpret the results in context, you won’t be providing the best possible service – and you’ll be depriving yourself of the chance to learn and progress.

How does in-house testing contribute to better patient care?
First, we facilitate and improve patient care by providing results much faster – and for many more patients – than a central lab. We also participate in tumor boards and, as I mentioned, we have regular telephone conversations with clinicians. This level of interaction is not possible when sending out tests to a central lab. I have had experience communicating with central labs; usually, the people I spoke to were not trained pathologists and lacked the expertise to address my questions.

The difference in turnaround times is also hugely important. We are talking about patients for whom even a single day – let alone a week or two – can make a lifetime of difference. Often, oncologists ask for urgent results because their patients are deteriorating. Cancer doesn’t take weekends off. Whereas a central lab might take two weeks to provide those crucial results, we can be flexible and expedite delivery.

There’s also the question of potential sample loss. Sending FFPE tissue blocks to a central laboratory means taking on the risk of losing them. That can lead to tragedy because, often, one result is not enough; we want to confirm that result via another method. If we don’t have the block, we can’t do that – and, in the future, we can’t use those samples for further testing or for clinical research (another way in which we support our oncologist colleagues). That’s a loss for the patient, the pathology department, and the institution.

What would you most like to emphasize?
The number of targeted therapies is growing and, one day, the standard of care will include a genomic profile for every tumor sample to help to guide treatment decisions. If we want to train the next generation of pathologists to understand molecular pathology and cope with these novel demands, they must be exposed to the entire testing process to give them the necessary education and experience.

Wei Song is Director of the Clinical Genomic Laboratory at the Englander Institute for Precision Medicine, Assistant Professor of Pathology and Laboratory Medicine at Weill Cornell Medical College, and Assistant Attending Pathologist at the New York- Presbyterian Hospital, New York, USA.