May 2025—In-house or send out, small panel or large, ease of use, getting the most information possible from a sample and doing it fast—that was at the center of a March 17 conversation on next-generation sequencing, led by CAP TODAY publisher Bob McGonnagle. “More and more patients need this testing faster and faster,” Jeremy Segal, MD, PhD, of the University of Chicago, said. What he and four industry executives told us in the online roundtable follows.
CAP TODAY’s guide to next-generation sequencing systems starts here.
In our next-generation sequencing systems roundtables, we often discuss the usefulness of bringing testing in-house, for the convenience of turnaround time and sharing results in the best way possible for clinicians, researchers, and the laboratory. The decision to bring NGS testing into an institution is a big commitment for those that might be concerned about funding and personnel, for example. Jeremy Segal, does some of the uncertainty affecting laboratory decisions mean that people who might otherwise have taken the big step to bring testing in-house might want to pause to see where they are?
Jeremy Segal, MD, PhD, director of molecular and cytogenetic pathology and professor of pathology, University of Chicago: It’s a mixed bag. We’re going to be squeezed financially in many ways and it won’t be limited to academic centers. If Medicare reimbursement goes down, private labs are going to be squeezed as well and will start asking for more payments. They already do. Some people feel you can send these tests to a private lab and you don’t have to pay and the patient doesn’t have to pay. That has been at times the way it works but not anymore. If it’s classified as an inpatient or within two weeks of an inpatient, private labs are charging hefty fees back to academic centers for those tests. That will only increase as private labs begin facing more pressure as well.
Our lab does the work at an overall lower price point than if we had to send it out. It’s about how you look at the finances and economics of it. It’s expensive to set up a program in-house, but if you’re sending everything out, you may end up paying more. The testing needs to be done one way or the other, and some places may choose to scale back validation while others try to accelerate.
A majority of people who are bringing NGS in-house also have great research interests, interests in publishing and improving scientific understanding. Robert Yamulla, how do you see the current appetite for research applications?
Robert Yamulla, PhD, MBA, research applications marketing lead, Illumina: On the research side, I’m looking at where we could go and what the future looks like 10, 15 years from now. What methods are being developed in the research space and moving toward a future in the clinic.
From a company perspective, even in the research space and especially in a limited funding environment, what matters is making sure you are getting enough information and interpreting the sample correctly the first time. We’ve been looking at how we can ensure quality in the end-to-end workflow, from getting the sample to preparing, sequencing, and analyzing it. Thinking of it holistically introduces value, scalability, and flexibility to many workflows because you can think of it as ensuring quality throughout the workflow.
Furthermore, how do we get more information from precious samples so we’re not missing something important? That’s where you start to hear discussions around multiomics. Getting to the right answer as quickly as possible involves getting additional layers of information, which you might miss by looking at one layer. The greatest movement we’ve seen is toward more information from a sample and more holistic views of a given system.
José Luis Costa, can you comment on this line of questioning?
José Luis Costa, PhD, global director of scientific affairs, clinical next-generation sequencing and oncology, Thermo Fisher Scientific: Bringing testing in-house is key but just as important a consideration is turnaround time. It can still take up to a month for in-house NGS to deliver results, and that’s not acceptable. By bringing this testing in-house, organizations should see improved speed on the result. Also, when you test in-house you have the flexibility to maximize what you can squeeze from the sample. This can help deliver high sample success rates so you get an answer in a shorter turnaround time for any sample in your workflow. That is something you can control in-house but is much harder to influence if you send out the sample.
The multiomics wave is coming but it’s still primarily focused in the research world. Companies need to differentiate what is clinical and provide the best clinical solutions but still leave bandwidth to bring in innovation, research, translational aspects, and multiomics. It’s the wave we are just starting to surf.
Shu Boles, as you strategize around genomics for Qiagen, you have a foot in both camps—research and clinical applications. What is your take on what you’ve heard so far?
Shu Boles, PhD, MBA, director of genomics strategic marketing, Qiagen: People are excited that we’re entering an era where we can generate so many layers of information from precious samples. It may not be directly leading to a diagnostic tool per se, but think about all the other possibilities, like drug discoveries or interventions, to potentially be part of patient care. Artificial intelligence is also a big topic. I’m excited for what we can do by combining these genetic and genomics tools to create diagnostic tools to discover different insights we can provide for better patient care.
We have considerable technology that helps patients, particularly cancer patients and many others, and yet if you go to the ASCO meeting and others, you’ll see studies that report fewer than 50 percent of cancer patients in a given cancer have the proper follow-up testing for biomarkers. Jeremy, do you have thoughts on a better solution to that problem?
Dr. Segal (University of Chicago): We don’t tend to see that here, and people at academic centers don’t necessarily see a lack of biomarker testing, so it can be hard for us to comment about. You hear about patients not getting tested mostly in community practice. It’s especially troubling when you hear about patients with EGFR-mutated lung cancer being put directly on immunotherapy without being tested, or other missed obvious biomarkers. Fixing it will really depend on educational programs targeting oncologists. It’s hard for pathology to meaningfully intervene unless there is an in-house lab and then the pathologists could focus on getting reflex testing set up. But if there’s no in-house lab, then pathology isn’t adequately involved and then the biomarker testing decisions are mainly in the hands of the oncologists. If there’s nobody helping on the pathology side and nobody forcing them to do it on the oncology side and it’s not part of their practice, maybe the payers need to get involved, though payers never seem to want to pay for our tests anyway.
Pierre Del Moral, what are your thoughts on this problem?
Pierre Del Moral, PhD, MBA, director of clinical marketing–oncology testing, genetic disease testing, reproductive health, Illumina: Eighty percent of cancer patients in general are treated in a community practice. Education is important. Effort is being made to educate community oncologists who have limited time and are seeing a variety of patients. Guideline testing is ever changing. It’s complex. We do see an appetite as well to internalize testing within community oncology practices. That education has made an impact. Now we’re seeing a shift toward the internalization of testing, even in community practices, and we will see changes in the number of patients tested as internalization becomes more prevalent in community practice.
José Costa, your comment on this question?
Dr. Costa (Thermo Fisher): The vast majority of patients receive care in community practices, so this is where testing should be done, which is a challenge. Raising awareness and spreading education can be a way to overcome some of the limitations, but there are also limitations in some technology. NGS platforms need to be simplified so they can be implemented in sites that have less expertise than academic centers, which often have robust expertise, bioinformatic pipelines, and infrastructure to support those platforms. In community settings, this technology needs to be tuned in a way that is more simplified for any laboratory technologist to run the test. We’re working to make the technology accessible.
I’m seeing the entrance of more platforms that are sort of way stations to next-generation sequencing. We’re talking about digital PCR or a panel that offers a limited number of mutations. The idea is use this first, you may find the mutation you are looking for. Shu Boles, can you comment on that?
Dr. Boles (Qiagen): We’re at a time where more information is power. That’s balanced with the turnaround time, the ease of use for bringing the test in-house. What is the right technology or combination of technologies to best serve cancer patients? Each lab or in-house testing site should decide for itself. I would be a strong proponent for helping our customers select the right technology so they can best serve their customers, whether it’s a digital PCR assay or NGS assay or a combination.
Jeremy Segal, can you comment on that in terms of rapid turnaround time options from your perspective?
Dr. Segal (University of Chicago): It’s one of the things that keeps us up at night. We have a big panel and a small panel. For certain tumor types the turnaround time requirements aren’t bad. The time pressure comes when results are needed for first-line treatment. With lung cancer, they want it fast. As things move from second or third line to first line, you see the turnaround time pressure go through the roof and more rapid options are needed.
Our large panel is great but it takes a long time, so we run a small panel on the side to give our oncologists information sooner. The problem is that the small panel isn’t totally comprehensive. It has some of the critical biomarkers but it doesn’t cover fusions. We do a FISH for ALK in a short time frame and some of the major biomarkers, but it’s not perfect. We’re looking at replacing the smaller panel with something more comprehensive, but many of the available rapid options aren’t ideal. They may find some fusions, some mutations, but not all. It’s hard to avoid the possibility of ending up in a pickle, where you gave a negative result with the preliminary panel but then the comprehensive panel results come out later and show a treatable event. If the patient was started on chemo, do you stop it and switch to something else? If the small rapid panel includes partner-agnostic fusion detection and all possible treatable biomarkers, then it starts to look like the slower comprehensive panel and it can’t necessarily be done in that rapid time frame. In the short term before we fully solve this problem, it requires close communication with our oncologists so they’re aware of all our results and the capabilities of our various tests.
Are you tempted to have a reflex testing algorithm for some cancers?
Dr. Segal (University of Chicago): We have it for lung and a growing list of other tumor types. As soon as the pathologists see a lung cancer they put the orders through for the small panel, our big panel, and our big RNA panel, and those results often come out before the patients have even seen their oncologist. We really like this model and we’re trying to expand it to as many tumor types as we can.
Dr. Yamulla (Illumina): Shu mentioned selecting the right technology to best serve the customer. One misconception, and you see it sometimes at conferences, is that people say, Will liquid biopsy replace H&E staining and sample? Will certain technologies circumvent another? Sometimes that’s the case, but sometimes it’s not. The type of information you get from these different modalities, resolutions, sample types, whether it’s a sample on a slide versus liquid biopsy, for instance, is sometimes complementary and sometimes additive. A liquid biopsy might be informative for what a tumor is releasing at stage three or four. Maybe it’s not as informative at stage one or two, and maybe it matters where cells exist spatially within a tumor. It matters where immune cells are infiltrating and where they’re not.
Dr. Segal (University of Chicago): I agree. For a long time there was a lot of “liquid biopsy is going to replace the tumor analysis.” I’m glad that’s kind of going away, but obviously plasma DNA testing does have a useful adjuvant utility: faster turnaround time sometimes, the ability to detect subclonality and survey a variety of different lesions in different places that you may or may not be aware of, resistance mutation detection, et cetera. I think patients should have all of the above, and then the question is, can you pay for all of the above?
Dr. Del Moral (Illumina): We recently saw changes on reimbursement with concurrent tissue and liquid biopsy coverage. So what Dr. Segal and Robert said is also happening on the payer side. It’s not happening fast enough, but at least there’s movement toward improvement in coverage, and we have the ability now to do tissue testing, the gold standard, and liquid biopsy to provide the best therapeutic option. We also see reflexive testing with pathologists initiating testing for indications agreed upon with treating oncologists. That helps with turnaround time and still achieves a comprehensive type of results. All this combined bodes well for better patient outcomes.
We’re now seeing a class of drugs called ADCs [antibody-drug conjugates], and they’re quickly becoming approved for pan-tumor applications, although the pan-tumor approval is dependent on the particular characteristics of the individual cancer for which they are approved. That can get confusing for some people. Pierre, does that complicate your life from an NGS point of view? That is, you need an NGS panel for the specific types of cancers that might be treated by the new compound.
Dr. Del Moral (Illumina): There are five or six pan-tumor or pan-cancer biomarkers—RET, NTRK, BRAF, HER2, and TMB are examples. TMB is probably controversial but that requires a large panel. RET and NTRK are fusions. Dr. Segal mentioned being able to capture those in a partner-agnostic fusion way, especially NTRK. Having the flexibility and scalability of the technology but also the panels, which can address rapid or comprehensive testing, allows us to cover these. As more pan-tumor companion diagnostics or indications are approved, the pendulum will shift toward larger panels.
I don’t believe it makes it harder on us. The theme of our discussion today is what is the best test, best technology, and the complementarity of the technologies used to address a given question. It depends on the treating physician, whether or not they’re looking at finding a biomarker that is more of a pan-tumor to make treatment available to patients, or if they want to look into the spectrum of a potential biomarker they’d like to identify to treat their patients.
José Costa, can you comment on this question of not only pan-tumor markers but also pan-tumor-targeted therapies?
Dr. Costa (Thermo Fisher): It’s not a challenge for assay design if it’s pan-tumor or not. ADCs are slightly different when compared with targeted therapies. The way we see biomarkers is different for targeted therapies—there’s either a mutation or a fusion gene or something that is more detectable, even if it’s pan-cancer. In the realm of ADCs, that is shifting what the biomarker might be, if there are biomarkers for identifying patients who would be better served with a specific ADC. There might be a space for research to understand how to better stratify patients for each ADC being developed. Still, ADCs are here and there are many clinical trials on ADCs, so the space will mature quickly to find ways to properly address patient stratification. ADCs still carry a toxic load on the ADC itself, so it will be important to make those assessments as precise as possible to best serve the patient.
Robert Yamulla, in the next year what do you see as your biggest challenges in your job and for your company?
Dr. Yamulla (Illumina): One big challenge is figuring out the right mix of technologies, methodologies, to characterize how to stratify patients. How do you determine ways in which different interventions could be most effective? How do you monitor their efficacy? Determine relapse? Determine next steps? There will be a combination of methodologies to be used in the research space to determine how you stratify, determine risk factors, and monitor over time.
The challenge is figuring out what to implement in your research and when. For instance, how do you stratify who will or will not respond to ADCs? There are many ways you can do that. I focus on multiomics, single cell, spatial technologies. Figuring out how the diverse tumor microenvironments respond to biomarker-specific therapies will be a huge area of research and a major challenge, but it’s also an area for opportunity. It’s going to involve a lot of research, understanding that these tumor microenvironments are incredibly diverse and even patient specific but doing our best to determine patterns by looking spatially or in single cells to monitor how these tumors evolve. What we’re trying to adapt these technologies to in research is how we deal with a system that continuously evolves. We’re not just looking at cancers that are homogeneous at one point in time. It is essentially Darwinian evolution gone awry. Figuring out how to monitor that using emerging technologies—like multiomics, single cell, spatial—will probably be a huge area of discussion over the next few years.
Shu Boles, what will be your biggest challenge in the coming year?
Dr. Boles (Qiagen): To help drive bringing assays in-house, whether it’s next-generation sequencing assays or digital PCR assays, to support the continuum of a patient’s cancer journey, from initial diagnostic to monitoring residual disease and response. Also to enable our customers, the testing labs, to adopt the right assay and then help get it up and running in the midst of a wide selection of new or developing technology that can enable clinics and generate research insights or drug development.
José Costa, same question.
Dr. Costa (Thermo Fisher): More than 50 percent of patients are lost during their diagnostic journey. We know the issues—they are not biopsied, the biopsy is not enough to be tested, the turnaround time is not fast enough to deliver results that inform care decisions. We have tools to overcome those, so how do we put those into practice more broadly? How do we change the paradigm so every patient can be provided with the proper biomarker testing?
The second challenge is the wave of multiomics. How do we integrate all this information? How do we bring genomics, proteomics, epigenomics, spatial AI into this equation? How can we put these technologies together in a way that it becomes decentralized in the lab to properly address clinical questions? With the breadth of technologies we offer, from early discovery and research to clinical diagnostics, this is a challenge we’re working to navigate alongside the industry.
Pierre Del Moral, what are your and your company’s challenges with the role of AI in your daily work?
Dr. Del Moral (Illumina): It’s a challenge and an opportunity at the same time. How do you integrate the most appropriate technology and results throughout the cancer patient testing journey and make the most impact? How do you gain the most amount of relevant information at the lowest cost while enabling the treating physicians and pathologists testing those samples to keep up with the technology? It is leveraging artificial intelligence to make it streamlined and more accessible and making the rationale for the testing so we can provide it to as many patients as possible and to the treating physician. The challenge is the integration of all this information and how we distill it in the most meaningful way to treat patients.
Jeremy Segal, what is the biggest challenge you foresee for you and the University of Chicago labs you run in the next year or so?
Dr. Segal (University of Chicago): A big challenge now is the emphasis on turnaround time. More and more patients need this testing faster and faster. We’re spending time in meetings with different groups trying to shave time off all the little parts of it. I’ll be happy if we get that worked out.
The bigger challenge is how we position ourselves for a complicated and exciting future. Multiomics, spatial genomics, microbiome—much of it is still research oriented, but every research project is one paper away from being clinically actionable, and then people start knocking on your door and asking when you can do x, y, and z. It’s not a question of trying to set up a spatial genomics clinical test today. It’s more about how we make sure we’re ready for it when it needs to be done and figuring how to get exposure to these areas and get early validation experience with it, so when the time comes we’re able to adopt it quickly.
When we first started, it was all new. Everything was development because we didn’t have a service. Now we have a service with people running clinical tests all day long, and now it gets complicated and becomes more difficult to bring on the newer technologies. I think about that all the time. How do we maintain nimble capabilities while growing into a mature and busy pathology service where much of our time is focused on process improvement, turnaround time, preauthorization, and billing. Half my day is billing and the other half is turnaround time, and it shouldn’t be. I have to figure out how we can carve out time to think about and plan for the future of genomics, cancer care, and research.