Precision Oncology Genomic Profiling: In-House or Centralized?

Targeted and immuno-oncology therapies requiring biomarker testing have proliferated in the last decade. Healthcare providers now face a complex decision: whether to outsource this new testing to centralized laboratories or implement it in their own labs. What is best for the system – and what is best for patients?

To explore these questions, we invited expert pathologists Philip Bennett from the UK, Fernando López-Ríos from Spain, Ruthy Shaco-Levy from Israel, and Michael Vieth from Germany to a panel discussion to share their views and experiences. They’ve participated in broadly different approaches to the issue, from limited testing hubs within a single country to plans to completely outsource testing to overseas commercial labs.

The discussion was moderated by Michael Schubert, editor of The Pathologist, and Luca Quagliata, head of medical affairs for Thermo Fisher Scientific’s clinical sequencing division.

What does in-house testing mean to you when it comes to oncology? 

RS-L: Performing all the pathology assays in my lab, from the hematoxylin and eosin (H&E) stain to the immunohistochemistry and molecular tests, is what appeals to me about in-house testing. It means correlating the molecular analysis with other clinicopathologic features to see the whole picture. For example, in breast cancer, we report on tumor size, tumor grade, and receptor status. For consistency and efficiency, molecular pathology should be included in that report – and performed on site.

FL-R: In-house testing means controlling the whole testing workflow so that you can influence turnaround times and other critical factors. In my opinion, it also enables us to put patients at the heart of the care process.

What is your institution’s approach to precision oncology testing?

RS-L: In my department, we perform all our molecular assays locally, because it benefits everyone involved. Turnaround times are shorter, there’s no need to send out precious samples, and clinicians can directly discuss test results with pathologists. The pathologists get to work with advanced technologies and fully develop their professional skills. All parties appreciate the high-level pathology reports with clinicopathologic correlation.

MV: Our system is driven by clinical and patient needs and follows a basic rule: all tests that can be performed locally should be.  If we encounter any problems with testing, a nearby university hospital can help us, but we try to carry out all routine tests in-house so that we build the expertise to handle not only simple, but also more complex cases.

What are the pros and cons of a centralized test model versus locally conducted testing?

PB: Centralized testing makes sense, for example, with a homogenous liquid biopsy or for certain biomarkers that are too rare to implement cost-effectively in every local laboratory. Unfortunately, some samples are sent out to hub laboratories with potentially lengthy turnaround times (and lacking any quality preanalytical assessment) just to get the standard-of-care biomarkers. This is a waste of resources. We must focus on doing those routine tests quickly, cost-effectively, and as locally as possible.

FL-R: With the advent of NGS panels, genomic profiling has become more leaner, cheaper, and more user-friendly. Everything is quicker in-house, with much less chance of losing important material or information. One of the best arguments for in-house genomic profiling is the control it affords over the preanalytical parameters, tissue specimen selection, and sample quantity.

MV: I also see an ethical issue with sending samples to commercial laboratories abroad. In Germany, the healthcare system is over 90 percent publicly financed and, if you spend this money outside the system in which it was generated, you aren’t supporting it and enabling its development – and this is an ethical problem.

Do you see value in increasing local knowledge and expertise in molecular testing?

RS-L: Yes. Pathology is one of the fastest-developing fields in medicine and molecular pathology is one of the fastest-developing areas in pathology. Soon, molecular pathology will likely be routine for confirming the diagnosis and prognosis of most tumors. Pathology departments not using these techniques will be left behind, so pathologists must develop expertise with the new testing methods – and with molecular pathology in general.

Have you experienced a move toward test centralization in your country?

RS-L: My hospital is part of a chain of institutions, and a few years ago, the decision was made to centralize our molecular pathology. The new central laboratory was not equipped to handle our testing needs, nor was it connected to our pathology department. Clinicians were not happy with the results or the fact that they could not properly discuss tests with the pathologists who had performed them. Eventually, the complaints mounted and the centralization attempt failed.

While this effort was underway, a wide gap developed between my hospital chain’s capabilities and those of hospitals whose labs had not been centralized and it took us some time to catch up.

How can in-house testing benefit your interaction with your colleagues – for example, in multidisciplinary teams coordinating oncology patient care?

FL-R: When we started NGS, we set up an internal “intra-laboratory molecular tumor board” to discuss test results before reports are released and clinicians and patients apprised of the results. It’s a formal meeting among the molecular biologists, pathologists, and technicians; we integrate the pathology information and individual biomarker testing with the NGS results and make sense out of the huge amount of information. This facilitates efficient conversation with not only our clinical colleagues, but also our patients, enabling them to understand and get the best value from their test results. It worries me that some people treat NGS results like something as simple and straightforward as cholesterol testing and think it’s enough to just send results to clinicians. These are complex tests with a lot of information that must be interpreted and put into context for every patient’s clinical situation.

Reducing turnaround time (TAT) to result is a hallmark of in-house testing. How important is it?

PB: No clinician ever complained about high-quality results arriving too quickly for any test, oncological or otherwise. It clearly impacts patient care. But it is also important to understand that the speed at which your lab can operate is not the only factor influencing TAT. Under General Data Protection Regulations, if you deal with cross-institutional or cross-IT systems, you are likely to encounter test result delays. The same applies to transporting samples.

What’s the value of keeping samples at your institution?

PB: This is the ideal scenario, and one of the problems with planned centralization is that people do not want to send out tissue blocks. However, if you outsource sections, curls, or slides from those blocks, you may be wasting material and not meeting preanalytical or sample requirement needs. Kept in-house, we can ensure that testers take only what they need from each sample. 

RS-L: It is important to preserve as much as possible of precious patient samples. If you do your testing in-house, you can decide on the test flexibly based on amount of sample available. The centralized labs perform the same large (over 500 gene) panels on all samples and sometimes do not get any results because there was just not enough of the tumor material. This means possible re-biopsy for patients and further delays.

Can any pathology laboratory today do genomic profiling for key predictive markers?

PB: From a technological point of view, we are near. The latest developments in PCR and NGS equipment are very much “sample in, result out.”  However, it will differ by country depending on the healthcare model. In the UK, following the 100,000 Genomes Project, there is substantial movement toward a few centralized molecular pathology laboratories. Some regional laboratories with existing skills and sample volume, like ours, are trying to stay in the game, but new laboratories will currently struggle to establish molecular techniques. 

FL-R: I agree. From a technical perspective, I can imagine that molecular profiling by efficient, actionable NGS panels will be relatively easy for most laboratories within a few years. But it all depends on how health systems organize their workflows. Currently, in Spain, most institutions have their own budgets and make their own decisions, but it’s a very mixed picture.

RS-L: In Israel, currently most, even small labs can do molecular tests with simpler methods such as PCR, FISH, or NGS assays. Larger academic hospitals perform NGS. I think in future, NGS will likely become even more routine and will be done even in smaller hospitals, because it makes sense for the clinicians and their patients as well as for pathology labs.

MV: With the recent advances in techniques and technologies, most pathology labs can certainly do NGS. It has to be cost-effective, of course, and you need qualified personnel – although not necessarily bioinformaticians these days.

Do you have a take-home message to share?

FL-R: I’d like to advocate for seeing things from the patient’s perspective. When we offer patients an NGS test, we also offer to discuss it with them. That tends to reassure them, because they value our honesty about the pros and cons of different treatment options, expectations, and possible problems. Ultimately, we need a patient-centered system, and that can only be achieved if we keep molecular profiling in-house.